platform

Voicegain Enterprise Speech-to-Text Platform (API, Portal, etc.)

Project maintained by voicegain Hosted on GitHub Pages — Theme by mattgraham

Release 1.118.0

New or changed functionality in the Transcribe App:

BE-3634 TA: Added call duration in header on transcript details page
QA-2236 TA: Disable LLM Playground box while the transcript is playing
QA-2461 TA: Improved role filter in Project Settings to display all options when any filter is selected for easier selection

New or changed functionality in other platform components:

BE-3482 Add epochReportedMsec to k8sStatus field on the onPremCluster API
BE-3626 Added llm field to question in the Call Review Config
BE-3577 Allow use of normal JWT token in GET /sa/call/search/fields and POST /sa/call/search
BE-3031 Copilot: Added ability to pull up notes after internal transfer.
BE-3628 Demo Voicebot: Display payment related fields in demo copilot
BE-3564 In AIVR App added openAiRealtimeApi as value of logicConnectMethod
BE-3526 In AIVR Session, log Error events for bad Callback responses
BE-3566 In POST /aivr add handling of VACA init if openAiRealtimeApi is value of logicConnectMethod
BE-3571 Modify mod_vg_tap to work without the rex_sid parameter (so that it can open socket to VACA)
BE-3621 Most read-only sql queries can now use the replica DB instead of the master.
QA-2464 Prevent the default JWT token for a Context from deletion via the Web UI
BE-3641 Return only 6 digits after the decimal point for confidence values
QA-2465 SA: Added option to delete Criteria in project configuration under project setting.
BE-3584 SA: Added trendlines to Yearly call trends chart on call stats dash
QA-2391 SA: Changed delete option from text to button on users settings page
BE-3586 Sa: Changed the zoom of the Yearly call trends chart to be along the x-axis only on the call stats dash
BE-3637 SA: Changed times in Call Sections card from timestamps to time durations on Call Details page
QA-2388 SA: Improved column spacing on user settings page to better fit longer fields like email
BE-3558 SA: Improved dark mode styling for Call Stats Dashboard
BE-3458 SA: Improved Design for ‘Confirm Configuration Changes’ Dialog on project configuration page
BE-3585 SA: Improved look of the charts on the Call Stats Dain in dark mode
BE-3617 SA: Improved performance of advanced search page
BE-3582 SA: Larger fonts used in Call Duration Statistics card on stats dash
BE-3593 SA: Make QA Call Review form visible
BE-3654 SA: Removed call selection checkboxes from all calls table to improve UX
BE-3583 SA: Removed extra gap on x-axis and increased line thickness in Call Duration Statistics card on Stats Dash
BE-3594 SA: Removed unnecessary scroll on advanced search page
QA-2469 SA: ‘Select All’ in advanced search filters now selects only the filtered results
BE-3624 Switch to using mod_shout for playback in AIVR
BE-3640 Web Console: Added 404 Error Page for api path which does not exist
QA-2446 Web Console: Added first and last name validations on user profile page
QA-2447 Web Console: Changed filtering ‘No rows’ placeholder text to “No Logs Data Found” in case of no match on Logs Page
QA-2410 Web Console: Copy button not working for SIP URI field in Phone apps drawer on Phone apps page
QA-2313 Web Console: Improved design of account management page to show all columns of login sessions section in a single view
QA-2168 Web Console: Improved language selection widget on the Settings Speech Recognition
QA-2431 Web Console: Improved placeholder messaging from “No rows” to “No Phone apps found” on phone apps page
QA-2414 Web Console: Increased font size of helper text for ‘Session ID’ and ‘Messages’ field on Logs Page for better visibility
BE-3559 Web Console: Replaced heavy background.svg to lighter png file
QA-2433 Web Console: Save button now remains disabled until the data cleanup interval is changed on settings page

Changes related to Integrity of Processing (fixes):

BE-3648 Fix a potential vulnerability where the session id for a new ASR session could be provided in the request
BE-3650 Fixed - Call Review Section name is not being saved
BE-3671 Fixed - GET /confgroup?type=SpeechAnalytics returns a plain JWT token instead of displayJwt
BE-3578 Fixed - GET /sa/call-stats does not seem to use the keywordGroup parameter
BE-3560 Fixed - StackOverflowException from a realtime SA session when no web socket connection is established for results
BE-3579 SA: Fix - Date and Time on stats dashboard should be in the format defined in the Project settings
BE-3580 SA: Fix - In stats dash, the duration display should support values larger than 59:59
QA-2466 SA: Fix - Negation toggle not working in Criterion Dialog in project configuration settings
QA-2389 SA: Fix - On API Security page, success message no longer shown without JWT creation. ‘Create’ button now requires valid name with note field validation
QA-2486 SA: Fix - Selected role not displayed in ‘Add User’ form on User Settings page after selection
QA-2224 SA: Fix - The end date of the demo project is earlier than the start date.
QA-2443 SA: Fix - The text inside call time breakdown card getting cut off on call details dash
BE-3581 SA: Fix - Total Call card values vertically aligned on call stats dash
BE-3627 SA: Fix issues in Call Review Form Configuration in settings
QA-1800 TA Edge: Fix - Meeting Minutes missing
QA-2468 TA: Fix - ‘Last Active’ column sorting not working in Project Settings
QA-2426 TA: Fix - Resolved overlapping ‘beta’ label issue in ‘Download as PDF’ pop-up on Transcribe Details page
QA-2427 TA: Fix - The Chinese flag displayed in the project settings is different from the project flag shown on the homepage
QA-2463 TA: Fix - The dropdown icon overlap with the exclamation (!) icon for the expected speakers field.
QA-2460 TA: Fix - The selected timezone is not visible in the timezone field after selection.
QA-2488 TA: Fix - Webex bot joins the meeting but doesn’t show as admitted on UI until the meeting ends where it then shows as admitted and failed to join.
QA-2365 TA: Fixed - The bot has already left the meeting but is still appearing in the meeting. This issue occurs for all three bots: Zoom, Teams, and Webex.
QA-2362 TA: Fixed - The meeting bot encounters a 500 Internal Server Error when attempting to leave a meeting.
QA-2072 TA: Fixed - The meeting bot is showing as admitted even though the meeting organizer has not admitted the bot.
QA-2364 TA: Fixed - The recording bot appears as admitted in the meeting, even though the organizer did not admit it.
BE-3638 Web Console: Fix - ‘Clear Filters’ button on Logs page did not clear Session ID and Messages fields
QA-2436 Web Console: Fix - Large extra space observed at the beginning of message when copied to clipboard and pasted on logs page
QA-2374 Web Console: Fix - Removed unnecessary scrollbar from the signup dialog
QA-2473 Web Console: Fix - Validation for numeric fields like ‘Max alternatives’, ‘Incomplete Timeout’ etc. on Speech Recognition Settings Page to not accept 0 as manual input
QA-2448 Web Console: Fixed - On the logs page, searching by message displays random errors and an internal service error in the console.

All changes affecting Security, Availability, Integrity of Processing, Confidentiality, Privacy are reported as such above. If nothing is reported in the specific category then it means there were no such relevant changes in this release.

Release 1.117.0

Key changes related to the core APIs

All builtin grammars are now available in Spanish version
Second dashboard for the Speech Analytics App
Errors in AIVR callback are now reported as events associated with a session
Beta version of Webex bot for the Transcribe App

New or changed functionality in the Transcribe App:

BE-3452 TA Edge: Pages for Webex OAuth handshake
BE-3530 TA: Add LLM Model to LLM settings
BE-3488 TA: Add support for Chinese language for transcription (beta)
BE-3496 TA: Added extra info to the Project delete dialog

New or changed functionality in other platform components:

BE-3471 Add a billingCode parameter to /asr/transcribe/async API
BE-3497 Add to onPremCluster API fields to define external Postgre
BE-3408 Added POST /sa/call/context/{contextId}/recompute
BE-3450 API methods needed for Webex OAuth hanshake
BE-3381 APIs to generate data for the generic dashboard on SA App
BE-3453 Bot that joins Webex meeting using SDK to collect speaker activity, audio, and video.
BE-3491 Implemented DELETE /sa/call
BE-3526 In AIVR Session, log Error events for bad Callback responses
BE-3527 In AIVR Session, log Error events for failed Callback requests
BE-3469 On Edge in influxDB measurements that have context - store context as a tag instead of a field
BE-3498 On Edge, all code using Postgres will use external Postgres if it is specified in onPremCluster Settings
BE-3464 Reopen transcribe-results websocket connection if connection is lost
QA-2338 SA/TA: Improve UI for Phrase detection config
QA-2342 SA/TA: Pre-set default value for the Location Channel under Phrases settings.
BE-3255 SA: Add a card that shows time distribution of various parts of the call
BE-3519 SA: Condense Agent Sentiment card to show 5+5 agents (5 best and 5 worst) in one view
BE-3332 SA: Generic Call Stats Dashboard
QA-2166 SA: Improved Keyword selection filter on Search Page
BE-3525 SA: Improvements in phrase settings
BE-3258 SA: In the Call Detail view on Debug page show the date when the call will expire
BE-3518 SA: In the Dashboard requests for call-stats change keywordGroup to “dissatisfied”
BE-3529 SA: Make smaller the icons in the top-right corner of the cards on the call detail page
QA-2275 SA: Make the toaster message for Add/Edit and Delete JWT disappear after 5-10 seconds.
BE-3512 SA: Swap Agents and Keyword cards on Dash
BE-3451 Store in the onPremCluster the UserId of the User who did Webex OAuth to obtain auth_token
BE-3349 Submit CRM notes at the end of Voicebot call (Claims Automation use case)
BE-3360 Universal HTTP webhook proxy via Cloud to Edge
BE-3499 Web Console: Add settings for specifying Internal/External Postgres
BE-3487 Web Console: Added Avatars
BE-3485 Web Console: On Edge view show Cluster Id
QA-2368 Web Console: Show Avatar of the Context creator on the page that lists all the Contexts

Changes related to Integrity of Processing (fixes):

BE-3457 Demo Voicebot: Fix - Start Over leaves the old polling still running if I make another call immediately
BE-3480 Fix builtin grammar for currency
BE-2835 Fix - API GET /share/{shareId} for a “within account” share, executed on a different account return 401 Forbidden, instead of 200 and empty response.
BE-3516 Fix - OOM from production asr-api
BE-3515 SA/TA: Fix - Rules for phrases too restrictive
QA-2363 SA/TA: Fix - The user is unable to update the Relation Parameter for the phrase group.
BE-3500 SA/TA: Fix bug with pop-confirm while deleting a phrase or phrase group
BE-3548 SA: Fix - Agent Dashboard - Top card’s text and line graph should be aligned to baseline
QA-2086 SA: Fix - Call Center ID filter has incorrect validation
QA-2398 SA: Fix - Overlapping issues on call overview page.
BE-3099 SA: Fix - PII redaction broken in /sa/offline
BE-3467 SA: Fix - Sentiment card keeps shrinking if there is not data in a chosen period
QA-2224 SA: Fix - The end date of the demo project is earlier than the start date.
BE-3513 SA: Fix - Top Dashboard cards look bad on a wide screen
QA-2399 SA: Fix - Unable to apply “Agent and Queue” filters in Advance search filter.
BE-3534 SA: Fix - Upload - Selected Queue name does not show on input field upon creation of new option
QA-1857 SA: Fix - User should not be allowed to create two projects with same name.
BE-3510 SA: Fix issues with Sentiment Card on the Dash
BE-3511 SA: Fix issues with the Topic Distribution card
BE-3505 SA: Fix the Call Time Breakdown
QA-1786 TA Edge: Fix - LLM Query is not working giving 500 Internal Server Error.
QA-2395 TA: Fix - getting something went wrong page for the first time login after signup to the application.
QA-2180 TA: Fix - LLM playground always returns summary in response to first question even if the question is not about a summary
QA-2339 TA: Fix - Phrase settings- the Min Word field should not accept negative or zero values.
QA-2397 TA: Fix - The toaster message ‘API token created successfully’ is incorrect, as the token is not actually created
QA-2343 TA: Fix - When a tag is separated by a comma, it should be added properly and should not remain visible in the input field after being added.
QA-2387 TA: Fix - When a user tries to create an API token without entering a name, the displayed message is incorrect. It should indicate that “Name is a required field.”
QA-2344 TA: Fix - When a user uploads a file with an invalid tag value, field validation should be applied to prevent submission.
QA-2369 TA: Fix - When User Click on Voice signature page than showing on page “speaker not found”.
QA-2359 Web Console: Fix - When the user clicks on ‘Transcription’ and uploads a file, the placeholder text for ‘Hint’ does not disappear after entering input.
QA-2360 Web Console: Fix - When the user clicks on ‘Transcription,’ uploads a file, and enters input in the ‘Hint’ field, the input text goes outside the box.

Release 1.116.0

Key changes related to the core APIs

Nuance SWI_ slots supported in MRCP NLSML return
Added DTMF tone redaction
Added ADDRESS to redaction options
Added recompute support to /sa/offline API

New or changed functionality in the Transcribe App:

BE-3382 TA: Implementation of a new design for Phrase Settings
QA-2203 TA: Improved error handling in the steps of Project creation
BE-3417 TA: Refactoring of Project Creation Flow and Project Settings for better separation of concerns
BE-3386 TA: Show the owner of the account in the user profile

New or changed functionality in other platform components:

BE-3399 Add option to our MRCP ASR to return NLSML in complete Nuance format
BE-3395 Add saAppHiddenFeatures to POST and PUT on the account.
BE-3362 Add to /asr/transcribe/async OFFLINE API ability to redact DTMF tones out of audio.
BE-3372 Added a method to apply redaction to existing /sa/offline transcription and audio
BE-3427 Added ADDRESS entity type to the redacting formatter (both transcribe and sa/offline)
BE-3408 Added POST /sa/call/context/{contextId}/recompute
BE-3407 Added POST /sa/offline/{saSessionId}/recompute
BE-3434 Added recomputePhase query parameter to GET /sa/call
BE-3426 Added two new read-only fields to sa/call – recomputePhase and lastRecomputeTime
BE-2920 Admin Tool: Add ability to control settings for Speech Analytics accounts
BE-2989 API for setting hidden features for SA Account
BE-3412 App Selector / Web Console: Inform about $5 credit for personal email
QA-2185 Demo Voicebot: Removed collecting user feedback
BE-3437 For Genesys platform support builtin grammars with languages other than English
BE-3279 In Webex Meeting Bot obtain email from the participants and store it
BE-3436 Made the language parameter for the builtin grammars case-insensitive and make it check only the language part of locale
BE-3322 Migrate Java services to spring-boot-3.2.11 (spring 6.1.16)
BE-3400 Remove asr-api as a dependency from sa-call-api
QA-2228 SA: Added a hover msg over Call Time Breakdown pie chart.
BE-3410 SA: Added CVV and ADDRESS PII Redaction type
BE-3432 SA: Added Recompute option to calls in the Call table
BE-3387 SA: Improve the time period selectors plus other Dashboard issues
QA-2270 SA: When searching for something not found in the transcript, now it displays 0/0.
QA-2020 SSO: Hide Forgot Password option on Edge
BE-3363 The artificial delay at Login API applies now only in case of non-200 response
QA-758 Updated messaging for App Selector
BE-3413 Web Console: Account signup - apply lower Credit to gmail and outlook.
QA-2268 Web Console: Add a back button on transcript page.
QA-2283 Web Console: Edit Phone App - A Close icon is provided to close the edit page.
QA-2306 Web Console: Meetings section displays ‘No Data Available’ instead of ‘No rows’ when there is no data for better user experience.
QA-2305 Web Console: Transcription section displays ‘No Data Available’ instead of ‘No rows’ when there is no data for better user experience.

Changes related to Integrity of Processing (fixes):

BE-3402 App Selector: Fix - Signup url takes me to login select page
BE-3447 Demo Voicebot: Fix - Bad Content Security Policy
QA-2333 Demo Voicebot: Fix - During the call, the voice assistant reports an error from the AIVR logic.
BE-3448 Demo Voicebot: Fix - Not loading the transcript (transcriptError)
QA-2246 Demo: Fix - file upload layout needs to be modified as the filename text is not properly aligned and appears cluttered.
BE-3438 Fix - /sa/call-stats is slow for the 30D period
BE-3371 Fix - If there are only 2 speakers, make sure that mergedAudio has 100% channel spearation
BE-3431 Fix - The vars returned in the Telephony bot DELETE callback are not stored
BE-3388 Fix recent vulnerabilities in 3rd-party Java Libraries
BE-3425 Fix: If using whisper model, Audio selector is ignored when creating stereo audio in offline-sa
BE-3391 SA: Fix - Do not show “+1 More” for tags, use that only with “+2 More” or higher
QA-2282 SA: Fix - Getting error while processing the uploaded file
BE-3445 SA: Fix - Sentiment weird for on certain screen sizes
QA-2266 SA: Fix - Settings -> Configurations : ‘Overtalk Percentage’ of ‘Incidents Threshold’ % value shouldn’t be more that 100%
QA-2265 SA: Fix - Settings -> Configurations : ‘Silence Percentage’ of ‘Incidents Threshold’ % value shouldn’t be more that 100%
QA-1811 SA: Fix -The user is logged out when deleting larger number of users at the same time from Users page.
QA-2237 TA: Fix - Input fields behave as if the user has entered special characters after adding a tag, as the tag requirements are being displayed.
QA-2259 TA: Fix - Latest News page, text formatting, background color and bullet point is missing.
QA-2336 TA: Fix - Meeting bot - not working for MS-Teams and Webex
QA-2116 TA: Fix - Project name should have a limit of 200 characters.
QA-2223 TA: Fix - Some participants’ details are hidden on mouse over when a large number of filters are selected
QA-2218 TA: Fix - The content on the zoom background light SVG image may appear confusing to users on the Zoom-App page
QA-2288 TA: Fix - Unable to Re-diarize transcripts for zoom upload directories.
QA-2331 TA: Fix - When the Regex Example tab is selected, the ‘Add Neural Network Example’ button should not be displayed, as it creates confusion about whether it is clickable.
QA-2102 TA: Fix - Zoom Meeting Bot is not recognizing that meeting has ended (for long meetings)
QA-2330 Web Console: Fix - “###” Extra text present in reslease notes section .
QA-2337 Web Console: Fix - Listen button clickable even though there is no saSessionId

QA-2284 Web Console: Fix - Telephony Bot API -> Business Configuration

‘Special Dates’ name shouldn’t accept the blank space

QA-2277 Web Console: Fix - Transcribe -> SA Configuration : ‘Overtalk %’ of ‘Incidents Threshold’ % value shouldn’t be more that 100%
QA-2276 Web Console: Fix - Transcribe -> SA Configuration : ‘Silence %’ of ‘Incidents Threshold’ % value shouldn’t be more that 100%
QA-2285 Web Console: Fix -Telephony Bot API -> Business Configuration ‘Named Prompts’ name shouldn’t accept the blank space

Release 1.115.0

Key changes related to the core APIs

Ability to switch language during async real-time transcribe session
Support for user-bound JWT tokens
Spanish sentiment and sentence similarity models
Text redaction supports chat and email formats directly
Outbound calling supports initial prompt
Voicebot Business Config supports multiple opening hours per day

New or changed functionality in the Transcribe App:

BE-3377 TA: Add In User Profile for Admin users ability to generate and manage user-bound JWT tokens
QA-2077 TA: Add success message when deleting a voice signature from account settings
BE-1468 TA: Add to Transcribe App Cloud an option to delete account
QA-2196 TA: Added input data validation for Webhook URL, Name, Token fields
QA-2143 TA: Change design for Keywords settings
QA-2225 TA: Voice signature - Play button now is disabled till the status turns into ready.

New or changed functionality in other platform components:

BE-3337 Add a field on aivr app to indicate a different fssk to launch outbound calls
BE-3350 Add ability to switch language during async real-time transcribe session
BE-3359 Add member information to the Call notes generated for the Provider calls
BE-3369 Add outboundPrompt to POST /aivr/dial/{destination}
BE-3373 Add outboundPrompt to POST /public-asr/aircall/user/dial-outbound
BE-3339 Add to POST /text/redact API ability to handle chat xml format
BE-3340 Add to POST /text/redact API ability to handle email html format
QA-2155 Admin Tool: Fix - Table Search- after clicking on the reset button table should reset to the previous value
BE-3376 APIs to create and manage JWT tokens that are bound to users
QA-1790 App Selector: Nicer formatting for Privacy Policy and Terms and Conditions pages
BE-3370 Copilot: Add outboundPrompt to the AIVR API dial command
MST-426 Deploy Spanish sentence similarity model
MST-406 Hang up the call when the user is calling a wrong line of business
BE-3329 Improve visibility into /sa/offline processing stats
BE-3358 In normalOpeningOurs deprecate open/close fields and replace with openPeriods
BE-3364 Limit Context name to 200 chars
BE-3325 Move /sa/call API to a separate microservice sa-call-api
MST-389 Multi-language design and framework in ml-svc
QA-2232 SA: Add success/alert notification after Add/Delete or Edit JWT
BE-3303 SA: Better Sentiment card on the Dashboard
BE-3320 SA: Better Topic distribution card
BE-3352 SA: Implementation of new Design for Phrase Settings
BE-3301 SA: Improved view for AHT card
BE-3302 SA: New design for time period selectors
BE-3374 Send email to support@voicegain.ai if an account status switches to “suspended”
QA-2020 SSO: Hide Forgot Password option on Edge
BE-2728 Support auto-reconnect if fssk loses connection to Freeswitch.
MST-431 Support MRCP confidence score scaling in rex
MST-430 Support multiple open/close prompt in business config
MST-370 Train and deploy the Spanish sentiment model
MST-407 Use LOB to filter account from the given phone number
QA-2186 Voicebot Demo: Remove Survey option.
BE-3375 Web Console: Provide info how to request Account cancellation and data deletion
QA-2114 Web Console: Add “Copy to Clipboard” tooltip on logs page
BE-3366 Web Console: Multiple open-close hours in the Normal Opening hours in the Business Config
BE-3218 Web Console: Redesign Home page (not logged in)
QA-2226 Web Console: The toggle button on SA Config is overlapping with the text.

Changes related to Integrity of Processing (fixes):

BE-3312 Fix - Customer reports that for each meeting they get 4 DONE webhook requests
QA-1992 SA: Fix - Inconsistent date format on call’s page.
QA-2119 SA: Fix - Project names should not be allowed to start with spaces or dots
BE-3379 TA: Fix - Project name entry cursor with strange behavior
QA-2092 TA: Fix - Add tag field is missing on the zoom directory upload page.
QA-2217 TA: Fix - Language dropdown is not centrally aligned during project creation
QA-2094 TA: Fix - Recomputing overlaping with size in storage.
QA-2253 TA: Fix - Settings- ‘Keyword’ name shouldn’t accept the blank space
QA-2252 TA: Fix - Settings- ‘Phrase’ name shouldn’t accept the blank space
QA-2073 TA: Fix - The tag field is not available when the user tries to save a browser share recording
QA-2220 TA: Fix - walkthrough wizard does not properly highlight features as expected
QA-2159 TA: Fix - When the owner downloads a transcript as a docs file, the transcript’s background color is not included in the file.
QA-2139 Web Console: Fix - Added Regex is not clearly visible in settings.
QA-2250 Web Console: Fix - API Security - ‘New Secret Token’ name shouldn’t accept the blank space
QA-2247 Web Console: Fix - ‘Download’ icon shouldn’t be active until transcribe status is ready like ‘View Transcript’ icon as if user click on
QA-2230 Web Console: Fix - For a brief moment, “Logout returned: Authentication token missing” is displayed.
QA-2249 Web Console: Fix - SA Configuration name shouldn’t accept the blank space
QA-2241 Web Console: Fix - Settings-> Speech Recognition: ‘Sensitivity’ should only accept value between 0-1
QA-2175 Web Console: Fix - Shortcut keys should be unique, and users should not be able to assign all shortcuts to a single key
QA-2147 Web Console: Fix - Unable to edit the description of any api security as it always shows the error
BE-3257 Web Console: Fix - various issues on the Logs page

Release 1.114.0

This release updates Edge deployments from Mongo version 6 to 7.

New or changed functionality in the Transcribe App:

QA-2183 TA: Advanced Search - Improve filter for selecting languages
QA-2190 TA: Advanced Search - Improve filter for selecting participants
QA-2189 TA: Advanced Search - Improve filter for selecting speakers
QA-2181 TA: Advanced Search - Improve filter for selecting tags
QA-2130 TA: The speakers table under account settings now has sorting functionality by Email.
BE-2874 TA: Introduced fallback error page for Loading Chunk Issue
QA-2137 TA: Show ‘Copy to clipboard’ message when hovering the mouse over the copy icon for the project ID
QA-2083 TA: Until a profile picture is updated by the user, display the avatar image in the profile settings
QA-2142 TA: When no group is available for the keyword to merge, display the option to create a new group

New or changed functionality in other platform components:

BE-3274 Add clientParams to openId Login API
BE-3273 Add languageDetection parameter to /sa/offline API
BE-3317 Add support for new CVV NER in our APIs
BE-2449 Added account API methods to add/remove Grafana support to account
BE-3311 Enable MRCP by default on all new Developer accounts
BE-3277 In Add Meeting Audio API method add email field to speaker or participant
BE-3280 In response from Login API return account Grafana setting if account has Grafana enabled
BE-3279 In Webex Meeting Bot obtain email from the participants and store it
BE-3090 Redesign the sso.voicegain.ai
QA-2198 Return from login API correct set of permissions for the user even if passwordChangeRequired is true
BE-3255 SA: Add a card that shows time distribution of various parts of the call
BE-3287 SA: Add a filter on Direction to the Calls table
BE-3266 SA: Add a set of fast filters for date selection to advanced search
QA-2053 SA: Add Save button for add/edit tags
BE-3306 SA: Better check for correct Keyword settings before saving them
QA-2184 SA: Better indication for case when sentiment has not been computed.
BE-3299 SA: Better read-only view of the AIVR integrations
BE-3298 SA: Increase size of Sentiment Icons
QA-2201 SA: Use the same rules on the call details page for considering an error as we use in the calls table
BE-3032 Support async writes to influxdb
BE-3292 Support CVV redaction in formatter in ASR API (transcribe, meeting and offline SA api)
QA-2099 Web Console: Copy to clipboard info shows when user mouse hover on copy icon
QA-2018 Web Console: Improved Edge logout behavior
QA-2006 Web Console: Improved transcript highlight efficiency so that pagination could be removed
QA-2082 Web Console: In user profile make clear that email field is read-only
BE-2448 Web Console: Only select accounts show Grafana option
QA-1867 Web Console: Show audio graph also for whisper models
MST-332 Train Kappa model using customer data
MST-357 Review and release the latest gpt-4o integration in llm-svc
MST-364 Improve billing intents routing
MST-380 Integrate new CVV feature in offline-main and ml-svc
MST-382 Support whisper-large-v3-turbo in APIs, and decide the default model for each use case
MST-409 Disable previous NER context when entering new context

Changes related to Integrity of Processing (fixes):

QA-2194 Fix for section formatting in Meeting API for certain cases
QA-2127 SA: Fix - Call Time Breakdown chart flickers when call is played.
QA-2234 SA: Fix - Date filter in Advanced search is not working when “Include current partial period” is un-selected.
QA-2165 SA: Fix - Keywords hidden on mouse over in advanced filter selection when a large number of keywords are selected
BE-3291 SA: Fix - Usage in Keywords must be optional
QA-2182 SA: Fix - Word cloud gets broken when user play call on overview page and then move to transcript page
QA-2138 SA: Fix - Word cloud toggle is missing in Confirm Configuration pop-up.
QA-1498 TA Edge: Fix - Unable to detect the Browser OS Device IP for all login sessions.
QA-2113 TA: Fix - Advanced Search: Project filter having extra white space that is covering whole page
QA-2129 TA: Fix - behavior of the tag entry box
QA-2110 TA: Fix - Duplicate tags shouldn’t be allowed
QA-2209 TA: Fix - Meeting Bot - Displaying ‘Forgot Meeting’ instead of ‘Leave Meeting’ for a Webex call.”
QA-2204 TA: Fix - Occasional unable to join Zoom meeting bot - Error: “Failed to join meeting”
QA-2120 TA: Fix - Project names should not be allowed to start with spaces
QA-2163 TA: Fix - Properly handle Projects with missing saConfig
QA-2160 TA: Fix - Unable to move multiple files across all projects, as the pop-up page displays differently and does not allow selecting the destination.
QA-2115 Web Console: Fix - New User Wizard: The header menu location is incorrect; it should be highlighted in the top right corner of the page.
QA-2013 Web Console: Fix - Alignment issue in the ‘Word Cloud’
QA-2122 Web Console: Fix - Column names e.g. Languages, Non-Real Time Models appear duplicated due to each ‘ASR Transcription’ and ‘ASR Recognition’ widget having an associated column
QA-2124 Web Console: Fix - Context filter should either have unique column names or provide a way to distinguish between the ‘ASR Transcription’ and ‘ASR Recognition’ widgets
QA-2216 Web Console: Fix - Edge Portal Logout is not working properly for Cloud Login.
QA-2105 Web Console: Fix - Fields names are different in table and column on call session page.
BE-3288 Web Console: Fix - In Phone number table, filtering by number does not work and/or is super slow
QA-1967 Web Console: Fix - ‘Level’ filter is not working as expected
QA-1953 Web Console: Fix - Log Filter is not working
QA-1404 Web Console: Fix - Logs filter is not working
QA-2170 Web Console: Fix - ‘Max Alternatives’ textbox of ‘Real-Time Acoustic Model’ section shouldn’t accept 0 as it’s value should be between 1-100
QA-1966 Web Console: Fix - ‘Services’ filter is not working as expected
QA-2097 Web Console: Fix - Sorting for Duration is not working properly on calls page.
QA-2098 Web Console: Fix - Sorting for Events is not working properly on call sessions page.
QA-2104 Web Console: Fix - Sorting is not working for ANI and FS UUID on call sessions page.

Release 1.113.0

This release prepares Edge deployments for Mongo upgrade from version 6 to 7.

This release reduces the scope of the permissions needed by pods that e.g. launch K8s tasks.

New or changed functionality in the Transcribe App:

BE-3226 TA: Add a title and heading to LLM Page.
BE-3229 TA: Disable invoices button if there are no invoices available
QA-2045 TA: When no speaker is present, the dropdown now displays the text ‘No speaker present’ under project creation

New or changed functionality in other platform components:

BE-3220 Add comfort noise to AIVR lua app
BE-3137 Added consecutive mode to GET /sa/call
BE-3240 ASR API now return urls with local host names and ports for internal requests
BE-3186 Populate fields in the markers field after a call is completed
BE-3188 Remove inactive login sessions faster
BE-3206 SA/TA: Better design for PII Redaction settings
BE-3255 SA: Add a card that shows time distribution of various parts of the call
QA-2052 SA: Add a search box for selecting the time zone under account settings.
BE-3215 SA: Add search on Time Zone selector
BE-3269 SA: Add sorting to the lists of values in Filters for Advanced Search
BE-3230 SA: Implement the call-section marking - audio line
BE-3258 SA: In the Call Detail view on Debug page show the date when the call will expire
BE-3253 SA: Mark sections of the call also on the timeline of the expanded audio view
BE-3233 SA: New icons for Inbound Call and Outbound Call
BE-3234 SA: On Call Detail Debug page show a link to the AIVR Session in Web Console
QA-1989 SA: On Calls page - improved data range selector
BE-3201 SA: Replace Topic Cloud with a pie/bar chart showing topics
BE-3283 SA: Support for phrase text queries in Advanced Search
BE-3237 Speed up the agent stats API
BE-3286 Web Console: Added clarification about formatter applicability to different transcription modes
QA-2004 Web Console: Added success message when creating a GREG Grammar
BE-3235 Web Console: Changes on AIVR session page: 1) obtain name from session, 2) add link to SA call details
BE-3243 Web Console: Improved ASR settings
BE-3146 Web Console: Improve look of the User profile page
BE-3254 Web Console: On the Context page make the languages the 1st column under ASR Transcription/Recognition grouping
QA-1982 Web Console: Remember selected column fields so that they survive page refresh
QA-2028 Web Console: Validate IP Address entered for IVR Proxy download
MST-350 Deploy the new Spacy NER model to replace the trained NN model in ml-svc and offline-main task
MST-352 Dockerize the latest RT ASR, REX and unimrcp

Changes related to Integrity of Processing (fixes):

BE-3250 Copilot: Fix - Refresh of the Pusher button does not work - but logout and login back works
BE-3252 Fix - Call-stats method sometimes fails with 500 error
QA-1999 SA: Fix - Call Center ID filter is not working in advanced search filters.
QA-1919 SA: Fix - Demo Calls are not generating in Demo calls in PROD.
BE-3214 SA: Fix - Gender and Word Cloud toggles do not work
BE-3212 SA: Fix - Mood selection does not work
BE-3259 SA: Fix - Sorting Calls by VB Transfer does not work
BE-3213 SA: Fix - Tooltip on NER behavior weird
QA-2011 TA: Fix - Downloading audio file with text file issue
QA-2046 TA: Fix - Microphone capture: The transcript scrollbar is overlapping with the start time.
QA-2064 TA: Fix - Zoom Meeting Bot is not working properly as it is failing even after joining a Zoom meeting.
BE-3261 TA: Fix speaker activity detection in Webex Meeting Bot if screen is being shared
QA-1952 Web Console: Fix - “Clear Filters” button is missing in Console log Filter section
BE-3185 Web Console: Fix - Collapsed LH menu icons are identical for many items
QA-2058 Web Console: Fix - Password Change button should disable after first click it should not invoke two calls
QA-1958 Web Console: Fix - Sorting of ‘Business Configuration’ is not working as expected
QA-2017 Web Console: Fix - Unable to change the speed of the transcript.
QA-2065 Web Console: Fix - Unable to download transcript under Transcribe+(beta) getting 403 (Forbidden) error.

Release 1.112.0

This release upgrades Mongo on Edge from version 5 to 6.

Key changes related to the core APIs

Improver Web Console UI.
/sa/call API supports call event marking which is utilized in the SA App to highlight various parts of the call
- NOTE: there is a known issue with the offset for the markers being somewhat off - this will be fixed in a patch
API for outbound calling that is used by Copilot
Copilot User may request Call Notes generation after internal transfer

Key changes related to Transcribe APP

New or changed functionality in the Transcribe App:

QA-1628 TA: Search functionality provided for Admin users when they delete any other user and transfer their projects to another user.
BE-2934 TA: Improved design of the Webhook Settings.

New or changed functionality in other platform components:

BE-3173 Add markers field to /sa/call
BE-3199 Added ignore_early_media parameter to POST /aivr/dial/{destination}
BE-2320 AIVR sessions will expire (using expiry period from AIVR App).
BE-3149 API that will be used to launch outbound call from the Copilot
BE-3197 Avoid storing null values under trace in each ivr_session document
BE-3204 Flag as error sa calls that will never recover from pending status
BE-3189 Implement the call-section marking
BE-3117 Improve performance of /sa/call-stats
BE-3202 In redaction, modify partial masking algorithm to always do a full mask if we have 4 or fewer digits total
BE-3161 Mark Meetings stuck in Processing as Error within 6 hours
BE-3109 Migrated Mongo 5 to Mongo 6 on Edge.
BE-3156 Modify account creation code to not create grafana org if type=SPEECH-WORKS
BE-3123 New API that can be invoked from Copilot that Agent can use to indicate transferred calls, so that Call Notes are generated and pushed.
BE-3021 Outbound Calling from Copilot with call submitted to SA App Project
BE-3186 Populate fields in the markers field after a call is completed
BE-3159 SA: Add “VB Transfer” column
BE-3203 SA: Add indicator for Fields data still loading
BE-3163 SA: Added a filter on Call Center Call Id
BE-3164 SA: Added more quick time period selectors for the calls
BE-3165 SA: Do not set sorting on the search query if query includes TxtSearchTerm
BE-3205 SA: Improved Tag editing
QA-1874 SA: More compact columns in the Calls table
BE-3154 SA: New design for the Keyword settings
BE-3125 SA: New SA project type - Generic Project
BE-3162 SA: On Call Details page show the time of the call including seconds
BE-3158 SA: Reduce the width of columns by shrinking displayed content and headers
BE-3201 SA: Replace Topic Cloud with a pie/bar chart showing topics
BE-3176 SA: Show playback position and total time in audio player
BE-3178 SA: Show Word Cloud
BE-3193 Store only top 100 words for WordCloud on the /sa/call
BE-2899 User records that have never been activated are actually deleted by DELETE api
BE-3184 Web Console: Added “Last Deployment Date” column to table of Edge deployments
BE-3198 Web Console: Added shortcut from app name in call session to the app definition
BE-3181 Web Console: Added UUID column to the table with Contexts
QA-1942 Web Console: Allowed User to enter time manually rather bound to select ‘Open’ and ‘Close’ time from clock selection in Business Configuration
QA-1988 Web Console: Better error messages in case of uploading bad grammar to GREG
BE-3216 Web Console: Improved Mode Switcher
QA-1984 Web Console: Improvements to Create GREG Question dialog
QA-1985 Web Console: Improvements to Create new GREG experiment dialog
QA-1616 Web Console: Move Transcribe+ from /sa API to /sa/offline API
BE-2997 Web Console: Redesign of customer portal.
QA-1998 Web Console: Switched Transcribe+ from /sa API to /sa/offline API
QA-1981 Web Console: Trim spaces from entries into login form

Changes related to Integrity of Processing (fixes): Issue key Custom field (Release Notes)

BE-3166 Fix - /sa/call/search when instructed to search NOTES seems to be searching the transcript text
BE-3168 Fix - aivrTransferDestType field on /sa/call does not seem to be populated
BE-3192 Fix - Cases where /sa/call is missing /sa/offline session
BE-3133 Fix - Hints in real-time word-for-word transcription mess up the results
BE-3222 Fix - redis memory leak caused by EventBus in Services that were not consuming events (only sending)
BE-3209 Fix - Unrecognized field “sessionDuration” during warm-transfer
BE-3170 SA: Fix - Error loop due to expired nonces
QA-1889 SA: Fix - In some cases call’s transcript only recognizes the agent and fails to identify the caller and their dialogues.
QA-2026 SA: Fix - Missing Tag Edit icon if call does not have any tags already
BE-3175 SA: Fix - Selectable items in search Filters should be show closer together and they should be sorted
BE-3183 SA: Fix - spacing between call id and search box
QA-1930 SA: Fix - The same number should not be allowed for both ANI and DNIS fields during call upload.
BE-3160 SA: Fix - Warm transfer audio (4 tracks) are incorrectly labeled
BE-3174 SA: Fix how the IDs are displayed on hover
QA-1931 TA: Fix - For accounts with OIDC enabled we should not show normal Signup page, instead we should show info about logging in with the OIDC creds.
QA-1950 Web Console: Fix - CLEAR FILTER button not clearing text from input fields in GREG Experiment Browser
BE-3180 Web Console: Fix - Context search is doing search also on date - it should search only Name
QA-1965 Web Console: Fix - Date range handling on the Logs page
QA-1949 Web Console: Fix - Experiment Browser Filter is not working
QA-1971 Web Console: Fix - If the imported GREG Experiment file does not match the required format, the import button should be disabled.
QA-1964 Web Console: Fix - Missing tooltips for the elements in the header
BE-3172 Web Console: Fix - Referrer Policy for https://console.ascalon.ai
QA-1972 Web Console: Fix - Search Reset behavior on User Management page
QA-2002 Web Console: Fix - Unable to upload Grammar file for creating Grammar under MRCP ASR.
BE-3157 Web Console: Fix AIVR App table width

Release 1.111.0

This release prepares for upgrade to Mongo 6 in 1.112.0

New or changed functionality in other platform components:

BE-3136 Copilot: Acknowledge all messages received via Pusher back to Voicegain.
BE-3130 Copilot: Added “Mark Call Transferred” button.
BE-3082 Implement mechanism to delete recordings from FreeSWITCH after they have been uploaded for processing
BE-3135 Implement POST /public/aircall/user/message-ack
BE-3086 Implemented trace field on the AIVR session.
BE-3123 New API that can be invoked from Copilot that Agent can use to indicate transferred calls, so that Call Notes are generated and pushed.
BE-3126 SA: Added the JWT token generation/management features as we have in Developer Web Console.
QA-1847 SA: Agents should not be allowed to edit PII Redaction in settings.
BE-2950 SA: Enhance the visibility of un-selectable rows for the user in both dark and light modes.
BE-3068 SA: Improve tag entry/edit - show existing tags already used within project.
BE-3125 SA: New SA project type - Generic Project
BE-3078 SA: Showing DTMF words in a distinct way in the Call audio timeline view.
BE-2863 Separate fav-icons for each product.
QA-1914 Web Console: Made Business Configuration UI user-friendly, so that users can manually enter the start and closing times.

Changes related to Integrity of Processing (fixes):

BE-3145 Fix - Hints with apostrophes should be allowed to be set on a context
BE-3114 SA: Fix - App may get into a loop
QA-1908 SA: Fix - Checkboxes are overlapping on advanced search filter pop-up.
BE-3035 SA: Fix - Credit card numbers are not being masked out in spite of NER redaction being turned on
QA-1787 SA: Fix - Demo calls are not generating for Demo projects.
QA-1909 SA: Fix - DNIS search is not working on calls page.
QA-1910 SA: Fix - Getting “Sorry, there was an error.” when using DNIS search.
QA-1822 SA: Fix - Password reset button were not responding.
BE-3113 SA: Fix - Sentry Integration for SA app as source-maps are not working for recent events.
QA-1809 SA: Fix - The agent created during the upload is not appearing on the Agents page.
QA-1813 SA: Fix - The owner is unable to edit their first and last name on the user page.
QA-1935 SA: Fix - The save button should stay disabled until changes are made on configuration page.
QA-1778 SA: Fix - Unable to access Demo calls getting “Sorry, there was an error! We were unable to download the call audio.
QA-1934 SA: Fix - Unable to edit project configuration name in settings.
QA-1927 SA: Fix - User is able to access integration page for Generic projects through left menu.
QA-1921 SA: Fix - When the keywords field is empty, it overlaps with ‘Average Handle Time’.
BE-2973 TA: Enhance the visibility of un-selectable rows (User table).
QA-1917 TA: Fix - ‘Continue’ button shouldn’t be enabled if project name is empty at first time login after signup.
QA-1901 TA: Fix - Delete Device text is not translated in German, Spanish and Hindi languages.
QA-1912 TA: Fix - Getting error while meeting Bot joining Webex meetings, “failed to join the meeting” even when the bots are present in the meeting.
QA-1906 TA: Fix - If a user tries to forget the meeting, the message ‘The Bot is not in the meeting’ should also be translated into languages other than English.
QA-1886 TA: Fix - Showing something went wrong page for the account setting.
QA-1913 Web Console: Fix - New User Wizard were not working properly.
BE-3131 Web Console: Fix - Removed hardcoded sensitivity setting from mic capture.
QA-1915 Web Console: Fix - When the user tries to explore the left menu for the telephone bot API, the business configuration text is overlapping and hidden.

Release 1.110.0

Key changes related to the core APIs

Multiple improvements and fixes to the Speech Analytics App

Key changes related to Transcribe APP

Improvements to the Webhook functionality
UI improvements

New or changed functionality in the Transcribe App:

BE-3075 TA: Added time-to-live (TTL) setting for the webhook JWT.
BE-2983 TA: Check the Error Boundary implementation and catch all errors
QA-1760 TA: Improved Profile page layout.
QA-1721 TA: Provide way to identify Zoom, MS Teams or Webex meeting.
QA-1821 TA: Removed Edit Tag functionality for the shared transcript for both public and account share pages.
QA-1851 TA: The password menu is available under the profile settings, so removed it from the “My Profile” menu.
BE-2940 TA: Updated Transcribe App styles.

New or changed functionality in other platform components:

BE-3074 Added jwtTTL parameter to Webhook API.
BE-2966 Allow Managers to get results of advanced search POST /sa/call/search and GET /sa/call/search/fields
BE-3101 Implement API to get all tags attached to /sa/calls in a given Context.
BE-3086 Implemented trace field on the AIVR session.
BE-2964 In the confGroup API throw Bad Request if someone tries to create a new confGroup without a name or with empty name.
BE-2986 Made SpeechAnalyticsCallDao optional so that asr-api can start without it in CHD environment.
BE-2978 Play audio while AIVR is making an outbound call triggered by warm-transfer.
BE-3052 SA: Added “Resend Invite” in menu under account Users.
BE-3039 SA: Added a cancel button in profile page.
BE-3088 SA: Added dailyRepeatCalls to API - /sa/calls
BE-3102 SA: Added filter for DNIS and a search for ANI.
BE-3073 SA: Added hours:minutes to the date filter in Advanced Search.
QA-1881 SA: In the login page, “Email Address” text is overlap on error message or pending approval message.
QA-1854 SA: Introduced option to delete “Examples”, “Keywords” and “Entity”.
BE-2938 SA: On Users table added filtering by Role.
QA-1766 SA: The status filter should function like the role filter, using predefined values instead of working like email filter on the Users page.
BE-3077 SA: Time-query to the back-end from the Advanced Search should use UTC.
BE-2863 Separate fav-icons for each product.
QA-1852 Web Console: Added maximum length limit validation for Business Configuration name.
QA-1827 Web Console: Added pop-up/ alert message appears when edit a user in User Management.
BE-2976 Web Console: In Phone number table, added a filter on Status column.
BE-2981 Web Console: In the AIVR App list, if the App has many phone numbers, then wrap them over multiple lines.
BE-3049 Web Console: In the Call Sessions page, DNIS column is now sorted in ascending order.
QA-1839 Web Console: Prevent Duplicate Shortcut Keys to Avoid Interference with Audio Playback.

Changes related to Integrity of Processing (fixes):

QA-1768 Admin Portal: Fix - Special characters and numbers are accepted in the “First Name” field.
BE-3096 Fix - Newly created Contexts from Web Console get type Transcription.
BE-3093 Fix - Syntax error in tsquery
BE-2944 Fix/Modify how we match names of meeting participants to the names obtained from Zoom API.
QA-1765 SA: Fix - Agents and Manager are able to access Users and Time setting page via URL manipulation.
QA-1802 SA: Fix - All demo calls are failing in processing and returning an error.
QA-1739 SA: Fix - ANI and DNIS fields in upload file should not accept characters as inputs.
QA-1840 SA: Fix - Call ID icon is overlapping with filter selector on Advanced search page.
QA-1888 SA: Fix - Call Time Breakdown in Call Overview is not showing for uploaded call with single channel.
QA-1855 SA: Fix - Calls tables column header and cell data are not aligned properly [1920 X 1080]
QA-1873 SA: Fix - Color inconsistency in delete project modal.
QA-1789 SA: Fix - Debug page is not clearly visible in Dark theme in call details.
QA-1850 SA: Fix - Direction filter shows is not working properly.
BE-3083 SA: Fix - Filter dates are reported in wrong format.
QA-1804 SA: Fix - PII Redaction feature not working.
BE-3100 SA: Fix - PII redaction in /sa/offline does not redact audio.
QA-1849 SA: Fix - Resolution field column is missing from calls table.
QA-1825 SA: Fix - The agent created during the upload is not appearing in the Agent filter on Advanced search page.
QA-1880 SA: Fix - The country flag icons are not showing up in the ANI and DNIS fields during call uploads.
BE-3095 SA: Fix - The date range component doesn’t work in Chrome.
QA-1857 SA: Fix - User should not be allowed to create two projects with same name.
BE-3006 SA: Fix - Weird chars on the Calls list pagination.
QA-1836 SA: Fix - Wrong pop-up msg on copying Voicebot ID.
BE-3103 TA Edge: Fix - Webhook page refresh is taking time and render page with no web-hook defined even there is existing web-hook.
QA-1791 TA: Fix - Cancel icon visibility is low in dark theme.
QA-1846 TA: Fix - “Title” field should be editable for Admin account user in profile.
QA-1780 TA: Fix - A type error page is showing on the advanced search page when the user tries to apply a filter.
QA-1826 TA: Fix - advanced search - The selected start and end dates are not populated in the UI, but they appear when saved in Date filter.
QA-1805 TA: Fix - Allowed domain settings are now visible for the Cloud version, whereas they were previously exclusive to Edge.
QA-1819 TA: Fix - An “unexpected character” error is displayed during signup on Transcribe Cloud
QA-1773 TA: Fix - An unexpected error page appears when a user tries to delete a device after changing its status to “Rejected.”
QA-1823 TA: Fix - Basic user should not be allowed to create share.
BE-2994 TA: Fix - Does not allow entry of valid hints.
QA-1812 TA: Fix - If the user is on the basic plan, the allowed email domain settings should not be displayed, as users on the basic plan cannot invite new users.
QA-1845 TA: Fix - In Dark theme input field text is not visible.
QA-1781 TA: Fix - Meeting Bot: If a user fails to join the meeting, they are unable to leave the meeting.
QA-1783 TA: Fix - Missing “Allow signup with the domain” field
QA-1833 TA: Fix - Partial translation when join a meeting by meeting bot.
QA-1797 TA: Fix - PDF Download: The meeting details table should be properly sized to fit the screen.
QA-1764 TA: Fix - Text was not translated based on selected language.
QA-1883 TA: Fix - The ‘Click here’ hyperlink appears on the login page immediately after entering the email.
QA-1856 TA: Fix - The user is able to add a Webhook without providing any details.
BE-3104 TA: Fix - Webhook jwtTTL information weren’t consistent.
QA-1869 Web Console [Edge]: Fix - Showing Transcribe App projects as context on the customer portal and user can also delete the projects.
QA-1700 Web Console: Fix - A vague error occurs when trying to upload files with long names.
QA-1870 Web Console: Fix - Creating a new context and switching to the other context shows a type error page on the customer portal.
QA-1859 Web Console: Fix - Creating a new context on the customer portal displays a type error and shows a blank page.
QA-1853 Web Console: Fix - ‘Open’ and ‘Close’ time can’t be same time of ‘Normal Opening Hours’ in Business Configuration.
BE-2980 Web Console: Fix - Sorting of the Phone Numbers by App name does not work.
QA-1897 Web Console: Fix - The Change Password page UI has layout and alignment issues.
QA-1882 Web Console: Fix - User is unable to select a time for “Open time” if “Close time” is not set
QA-1904 Web Console: Fix - When a user hover at billing label in header section, no cursor:pointer

Release 1.109.1

New or changed functionality in other platform components:

BE-3088 Add dailyRepeatCalls to /sa/calls
BE-3011 Add dtmfEvents field to /sa/call
BE-2968 Add fallback in case of Azure TTS issues
BE-2877 Added new variable in /sa/call that stores the type of transfer - Internal, Agency, None.
BE-3087 Additional fields added to the Pusher messages to be sent to Copilot
BE-2966 Allow Managers to get results of advanced search POST /sa/call/search and GET /sa/call/search/fields
BE-3065 API for text redaction POST /text/redact
BE-2961 API introduced to get Zoom Server API access_token.
BE-1820 Ascalon-cleanup to support removing expired sa_offline sessions.
BE-2957 At the end of the AIVR session, set language field of the /sa/call based on the language of the AIVR session.
BE-2959 Copilot: If not logged in show message about need to login.
BE-2974 Copilot: Implementation for updating co-pilot extension for Edge/Chrome browser.
BE-2985 Copilot: Improvements in settings page.
BE-3016 If a call came from a provider then prepend to the GPT Call Notes the info about provider from Voicebot
BE-3012 In /sa/offline obtained from /sa/call add DTMF values into the words returned
BE-2958 In POST /sa/offline/call/{callId} use the language specified in /sa/call unless overridden by settings.asr
BE-2944 Modify how we match names of meeting participants to the names obtained from Zoom API.
BE-3015 New API method to modify sa call tags
BE-2902 Poll URL returned from POST /asr/transcribe/async is a sticky url.
BE-3010 Record DTMF detected in AIVR session
BE-3023 SA: Add a direction column to call tables
BE-3040 SA: Add Regex formatter to the PII Redaction settings
BE-3064 SA: Add search functionality in Filters
BE-2945 SA: Added ANI and DNIS to the Call table and Advanced Search table
BE-2941 SA: Added callCenterCallId to /sa/call and set it from aivr session aircallId.
BE-2956 SA: Added language field to /sa/call.
QA-1730 SA: Added spinner when deleting admin user to show admin user is being deleted.
BE-2936 SA: Added status of the Users in the Users Table.
QA-1732 SA: Added user’s role in the profile section.
BE-3000 SA: Browser-specific text on the extension download page
BE-3057 SA: Change the number of calls retrieved on the Calls page to max 1000
BE-3022 SA: Consistent ANI/Dialed Number and CLI/DNIS naming
BE-2939 SA: If no sentiment value is returned from the API we are showing “N/A” along with icon/emoji.
QA-1751 SA: Improved design for Agents & Agents Detail page.
BE-2954 SA: In Call Details added a Debug page that is visible to the Admins.
BE-3070 SA: Increase the limit for number of calls in the Export query from 1000 to 2000
QA-1752 SA: Made company details in profile read-only for Manager
BE-2962 SA: New design introduced for Agent pages.
BE-3044 SA: Remember number of items per page setting till user modify.
BE-2999 SA: Show callCenterCallId in the Call in Calls list
BE-3013 SA: Show DTMF words in a distinct way in the Call transcript view
BE-3014 SA: Show tags and allow adding new or editing existing ones
BE-3027 SA: Show the next/prev buttons also on a call details page that is in processing right now
BE-2652 Sync AIVR session with Aircall callId at transfer
QA-1693 TA: If the number of speakers exceeds the expected amount, providing a scroll bar. This will not affect the transcript view.
BE-3017 Web Console: Add a refresh button on the AIVR Call Session Details page
BE-3004 Web Console: In AIVR session list, remove Version column, and add IVR Sid and FS UUID columns - both searchable
BE-3048 Web Console: In the Call Sessions page, On the ANI column, instead of a Selection Filter we will have Search option.

Changes related to Integrity of Processing (fixes):

QA-1734 Demo: Fix - For Long transcripts- text were overlapping with buttons.
QA-1735 Demo: Fix - Typography issue on the demo healthcare call details page.
BE-3043 Fix - Bad Event Bus configuration causing some events being dropped
BE-3041 Fix - Did not report Queue and Agent in /sa/call
BE-2991 Fix - Enforce non-empty name on Contexts in a correct way
BE-3045 Fix - GET /aivr-app/uuid/jwt fails to return JWT tokens using authToken in some cases
BE-2996 Fix - GET /asr/transcribe/uuid/transcript returns 500
BE-2987 Fix - Glitch in business operating hours/days
BE-3024 Fix - Issue with ascalon-asr-api starting on Edge deployment (specifically smopqa)
BE-2780 Fix - netty.io Buffer leak in Java FreeSWITCH ESL library
BE-3061 Fix - NPE when reading IvrSession.Recordings from Firestore
BE-2943 Fix - NullPointerException in Aircall webhook
BE-3058 Fix - Phone number formatter StringIndexOutOfBoundsException
BE-3003 Fix - PII redaction does not work for /sa/offline transcript
BE-3009 Fix - Real-time transcript possibly available too late for Call Notes
BE-1366 Fix - RedisConnectionException: Issues with Redis Marshalling Codec
BE-3020 Fix - Some Voicebot calls that get transferred end up in Disconnect instead of Hangup
BE-3060 Fix - START-INPUT-TIMERS is sent to Rex right after RECOGNITION-COMPLETE is received
BE-3050 Fix - The Users that we create when saving /sa/calls that have Agent data, do not have accountId
BE-3066 Fix - Voicebot to Aircall call transfer handshake error.
BE-2953 Fix - Voicegain call incorrectly matched to data from Aircall.
QA-1784 SA: Fix - Admin account user is able to change Owner account user’s role.
BE-3047 SA: Fix - Advance search is not working for ANI/Dialed Number
QA-1740 SA: Fix - After successfully deleting a user through the checkbox, another user gets selected.
QA-1761 SA: Fix - Agents and Manager should not delete project created by Admin.
BE-3018 SA: Fix - Download from Advanced Search does not use the same max calls setting as the search
BE-3037 SA: Fix - If we are using a 12 hour format we should show AM/PM in the time displayed
BE-3046 SA: Fix - Need to access next/previous call transcript, if there is no call audio available.
QA-1753 SA: Fix - No calls are showing on Advanced search page getting 500 (Internal Server Error).
QA-1769 SA: Fix - Placeholder text does not disappear when the user enters data.
QA-1831 SA: Fix - Sorting is not working for ANI and DNIS fields on Calls and Advanced search page.
QA-1698 SA: Fix - The Call’s Overview page is not responsive when zoomed in, causing text to overlap.
QA-1706 SA: Fix - The Queue option in the file upload under Integration displays duplicate entries.
BE-3076 SA: Fix - The search query in the Advanced Search Page should by default include sort by date so that most recent calls are retrieved first
QA-1738 SA: Fix - Unable to delete multiple users when there are admin users among them.
QA-1528 SA: Fix - When a user uploads a corrupted audio file, an error message repeatedly appears.
QA-1157 TA Edge: Fix - For some audio URL showing something went wrong issue.
QA-1800 TA Edge: Fix - Meeting Minutes missing
BE-2951 TA Edge: Fix - Projects returned from the back-end are ignored.
QA-1723 TA: Fix - “Meeting Minutes” string should not appear two times.
QA-1708 TA: Fix - [Dark Theme] When the owner attempts to rename the transcript, the cancel icon is not visible.
QA-1717 TA: Fix - After file uploading, the ‘File uploaded’ text weren’t translated.
QA-1757 TA: Fix - Edit transcript icon overlapping with the transcripts name.
QA-1701 TA: Fix - Hindi translation in “Delete Transcribe” pop-up.
QA-1707 TA: Fix - Made Phrases visible for the dark theme.
QA-1716 TA: Fix - Meeting Bot Placeholder text translation according to selected language.
QA-1709 TA: Fix - No space between “Timezone” text and the dropdown box border.
QA-1702 TA: Fix - Time zone pop-up was hiding the left menu.
QA-1724 TA: Fix - When owner unable to join meeting bot receive “Failed to join meeting” string was not translated as per selected language.
QA-1663 TA: Fix - When the Hindi language is selected, the ‘Apply’ button overlaps with the participants dropdown on the advanced search page.
BE-2984 Web Console: Fix - Timezone prompt not taking spaces as input
QA-1685 Web Console: Fix - When admin try to create “Add broadcast websocket” receive 400 bad request error.

Release 1.109.0

Key changes related to the core APIs

Added Webhook API (phase 1)
UI redesign for the Speech Analytics App
Speech Analytics app now supports Warm Transfer scenario
Copilot Browser Extension for Agent Assist

Key changes related to Transcribe APP

Support Webhook API (phase 1)
Added third set of LLM service settings and add model name field to all 3 LLM services.
Including transcript metadata together with transcript when using LLM to generate meeting minutes items.
Improved reliability of the Meeting Bot on longer meetings

New or changed functionality in the Transcribe App:

BE-2798 TA: Added new LLM Prompts for Meeting Minutes.
BE-2855 TA: Added third set of LLM service settings and add model name field to all 3 LLM services.
BE-2805 TA: Added user’s status column in users lists
BE-2860 TA: Added Webhook settings page.
BE-2857 TA: Addition to the LLM Prompts Settings.
QA-849 TA: Consistent password complexity requirements.
BE-2812 TA: Enforced LLM token limit.
BE-2850 TA: Including transcript metadata together with transcript when using LLM to generate meeting minutes items.
BE-2874 TA: Introduced fallback error page for Loading Chunk Issue.
BE-2875 TA: Introduced fallback page for error boundary case.
QA-1228 TA: Sorting given for users by name, for the Login all session table.

New or changed functionality in other platform components:

BE-2856 Add Korean (ko) as one other possible value of switchLanguage in AIVR callback response.
BE-1043 Add secure SIP and RTP to freeswitch.
BE-2450 Added ability to obtain temporary JWT token from AIVR session using authToken
BE-2891 Added metadata value to the speakers field in API /sa/offline
BE-2714 Added new Group of items sent to Aircall as Caller Insight.
BE-2843 Added originatingCallId and spawnedCalls to API - /sa/call
BE-2819 Added processing status to API - /sa/call
BE-2226 Added Webhook Notification APIs.
BE-2896 Adding in API - /sa/call an Agency tag, If AIVR inbound call is an Agency call (warm transfer bridged).
QA-1633 Admin Tool: Updated Terms and Conditions page.
BE-2832 API - /public/aircall/user-check used to check if user info matches data in Aircall.
BE-2904 Copilot: Create a web interface to host the .crx download links
BE-2884 Copilot: Implemented override of the Intent by Agent.
BE-2861 Copilot: Implemented the Login.
BE-2888 Copilot: Remove the dev/qa/prod selector and /dev/qa indicators.
BE-2847 Copilot: Show Intent card with Agency Transfer data.
BE-2876 Edge: If Action Items LLM prompt is not defined, then do not generate Action Items at all.
BE-2897 For real-time transcription sending metadata over results websocket, also every time metadata is added/modified.
BE-2886 Generate full call notes for the agency call that was bridged to the caller (not transferred to Aircall)
BE-2826 Handling 2x2 audio channels from warm transfer call in API - /sa/offline.
BE-2824 Improved design of Copilot.
BE-2825 In AIVR session allowing for caller id to be any of the secondary phone numbers attached to the AIVR App.
BE-2889 Included aivr_authToken and aivr_sessionId variables when pushing the AIVR Voicebot variables to the Copilot.
BE-2787 INVALID_ARGUMENT: Sample rate must be empty or 24khz for this voice.
BE-2929 Log the response payload from the Zoom API method that gets info about Zoom Meeting participants
BE-2320 Make AIVR Sessions expire (use expiry period from AIVR App)
BE-2928 Make sure that all calls in SA Demo project have progressPhase=DONE
BE-2775 make sure that user deletion does not fail if context deletion takes too much time
BE-2854 Modifications to llmSettings field (introduced premiumLlmService) on the onPrem cluster API.
BE-2710 New version of the Sidepanel (Copilot Browser Extension) that integrates with Aircall.
BE-2892 Populate speaker metadata in API - /sa/offline/call/{callId}
BE-2844 POST /sa/offline/call/{callId} need to be able handle calls with spawnedCalls (from warm transfer).
BE-1366 RedisConnectionException: Unable to init enough connections amount! Only 0 of 24 were initialized.
BE-2829 Return 400 Bad Request if someone tries to create a new AIVR App or assign a phone to existing App if the phone is not available.
QA-1695 SA : Highlighting any calls that had error in processing
BE-2885 SA: Added a Export button for the Advanced Search Results.
QA-1623 SA: Added limit of 1024 characters against description under General settings.
BE-1796 SA: App should be customisable per account.
BE-2808 SA: Enabled currency formatting for /sa/offline in AIVR sessions.
QA-1587 SA: Enhancement - Downloaded call transcript should have Call ID as name instead of timestamp.
BE-2933 SA: For call overview, render in processing page rather error page if call is in progress.
BE-2704 SA: For the call sessions that are being processed, adding a subtle overlay that says processing
QA-1695 SA: Highlighting any calls that had error in processing
QA-1533 SA: Optimized call page & made fields small enough to fit all data without scrolling.
QA-1690 SA: Provide correct information for Aircall Integration
BE-2818 SA: Redesign of Speech Analytics App UI
BE-2916 SA: Show first and last name in the user table
BE-2862 SA: Showing more than 2 channels in the SA Call Detail view if present.
BE-2907 SA: UI changes to implement the new design
BE-2879 Setting ANI correctly on calls coming from AIVR Voicebot to Aircall.
BE-2881 Submitting tags to Aircall when doing the call handshake between AIVR and Aircall.
BE-2728 support auto-reconnect if fssk loses connection to Freeswitch
BE-2836 Supporting Korean Prompt in telephony bot and in audio server.
BE-2883 Taking the Voicebot intent and insert it into /sa/offline and /sa/call.
BE-2880 Updated Advanced Search API - /sa/call/search to return results in CSV format.
BE-2859 Use premium LLM Service with the new LLM Query logic and use standard with the old LLM Query logic.
QA-1536 Web Console: Added popup after adding or deleting a user under User management.
QA-1540 Web Console: Added popup for deleting a WebSocket under Live Broadcasting.
BE-2001 Web Console: Notify about inability to to microphone capture on HTTP urls.
BE-2830 Web Console: Remember last items per page setting on the Call Sessions page.
BE-2831 Web Console: Remember the last value of items-per-page setting on the Phone Apps page.

Changes related to Integrity of Processing (fixes):

QA-1635 TA: Fix - “MINE/SHARED/ALL” option should not be visible to Basic account on LLM Query page.
QA-1225 TA: Fix - Clicking on backward icon of any transcript results in value of time getting negative.
QA-1665 TA: Fix - Clicking on microphone capture save recording page showing, it should show the start recording pop-up page.
BE-2842 TA: Fix - Dialog to reset password.
QA-1485 TA: Fix - Error page when opening rediarized transcript (partial fix)
BE-2906 TA: Fix - error when searching for a phrase with an apostrophe
QA-1604 TA: Fix - Footer of Document should not overlap with transcript in downloaded PDF.
QA-1689 TA: Fix - For longer meetings, after it ends, the transcript isn’t processing properly.
BE-2900 TA: Fix - If a Zoom meeting is long and is eventually ended by the host the bot does not notice that meeting has ended.
QA-1605 TA: Fix - In share meeting pop -up , month name is not translating and always show in default language.
QA-1497 TA: Fix - Invalid date and year showing for the last active under account users.
QA-1672 TA: Fix - LLM Query not working and returning 500 error (Internal Server Error).
QA-1718 TA: Fix - Meeting bot icon should also show as disabled when the maximum allowed minutes limit is exceeded.
QA-1490 TA: Fix - Meeting Bot were unable to join Webex meeting.
BE-2845 TA: Fix - Meeting platform selection tiles were too indistinct.
QA-1677 TA: Fix - Some old transcripts still showing the queued status and user can not delete or remove them.
BE-2195 TA: Fix - Stuck on login.
QA-1662 TA: Fix - The error message should be displayed in the selected language after the maximum allowed minutes have been exceeded.
QA-1667 TA: Fix - The walkthrough wizard is not functioning properly for new users.
QA-1600 TA: Fix - Unable to join the webex meeting by meeting bot.
BE-2834 TA: Fix - Unexpected Application Error! Loading chunk failed.
QA-1609 TA: Fix - Voicegain logo is not fully visible in the 100% zoom.
QA-1668 TA: Fix - When a user selects more than 12 speakers, the submit button becomes disabled however we have limit for 20 speakers.
QA-1560 TA: Fix - When user try to invite user, it shows an invalid email address error but when user re-click on the save button it showing “User already exists”.
BE-2809 Fix - Bad currency formatting for thousands plus
BE-2722 Fix - Bot logic doesn’t get input callback if the caller hangup immediately after saying something.
BE-2436 Fix - In meeting Join if the pod/puppeteer leaves the meeting because it is not progressing there is no info about that event
BE-2814 Fix - MS-Teams puppet bot were not terminating if the meeting gets ended.
BE-1354 Fix - NatsEndpoint fails to reconnect to NATS when its connection is lost
BE-2869 Fix - Not working error prompt in telephony-bot LUA
BE-2927 Fix - Pusher sending DNIS instead of the ANI
BE-2663 Fix - Several calls where the merged meeting audio is missing a speaker.
BE-2923 Fix Demo Prod
BE-2913 SA: Fix - Agent and Caller labels are incorrect
BE-2918 SA: Fix - Call Transcript page failing in case of single speakers
QA-1699 SA: Fix - Exports in advanced search is not working with filters.
QA-1686 SA: Fix - Getting error on Call’s Transcript page.
QA-1733 SA: Fix - Issues when Agent/Manager user try to edit/delete Admin user.
BE-2915 SA: Fix - New User Dialog starts with values of the previously created user
BE-2917 SA: Fix - Save button not activated for all values on Profile page
QA-1729 SA: Fix - Selected duration is not shown in applied filters in advanced search page.
BE-2898 SA: Fix - Spinning forever if the /sa/call has no saSesisonId.
BE-2898 SA: Fix - Spinning forever if the /sa/call has no saSesisonId.
QA-1719 SA: Fix - The added user role changes to “User” even if a different role was assigned while inviting.
QA-1725 SA: Fix - The calls continue to load and display an error icon when hovered over, but the call itself is still accessible.
QA-1622 SA: Fix - Timestamp interval dropdown have values that are more than transcript runtime.
BE-2914 SA: Fix - Unable to add Admin Users
QA-1710 SA: Fix - Unable to edit role of user under Profile users.
BE-2926 SA: Fix - User Edit dialog has just First Name and it does not even populate it
QA-1596 SA: Fix - User is unable to modify AIVR app getting “400 Bad Request”.
QA-1687 SA: Fix - User is unable to upload file under Integration section.
QA-1528 SA: Fix - When a user uploads a corrupted audio file, an error message repeatedly appears.
QA-1535 Web Console: Fix - Audio URL field box is too small.
QA-1625 Web Console: Fix - For Phone Management, clickable area is too small.
QA-1571 Web Console: Fix - Logs - When a user tries to filter logs, filter not working.
BE-2908 Web Console: Fix - Phone number deletion from AIVR App does not work
BE-2813 Web Console: Fix - Playback timeline cut off after the last word.
QA-1509 Web Console: Fix - ‘Speed and Audio’ control panels should be in proper size.
QA-1392 Web Console: Fix - When we click the back button more than once, the time interval goes to minus and after that the video pauses.

Release 1.108.0

Key changes related to the core APIs

More accurate offline transcription model
- Removed artifacts from training data
- Added additional Call Center data to training set
Finalized warm transfer support in Telephony Bot API
Support for pushing Voicebot data and Call Notes to Copilot

Key changes related to Transcribe APP

Cloud only: Major improvement to capabilities of LLM Query
Major makeover of the look of the UI
Added UI dark mode
Added Configurable Meeting Minutes (via LLM prompt)
Added tracking of LLM tokens (in the Cloud)
On Edge: Added ability to retrieve speaker data (incl. email) using Zoom API

New or changed functionality in the Transcribe App:

BE-2798 TA: Added new LLM Prompts for Meeting Minutes
BE-2811 TA: Added Purple project color (instead of black).
QA-1003 TA: Add option to rename transcripts.
BE-2668 TA: Change Key Items to Meeting Minutes : Add more categories in addition to Action Items
BE-2762 TA: Change naming from Key Items to Meeting Minutes.
BE-2786 TA: Changed border radius on avatars.
BE-2718 TA: Dark mode is enabled for transcribe app.
BE-2763 TA: On LLM Settings have two tabs - one for Prompts and one for Services
BE-2797 TA: Post signup to Transcribe App, sending signup email.
QA-1515 TA: Show creator avatar on the transcript detail page
BE-2781 TA: Show usage of LLM Tokens
BE-2789 TA: Showing email icon next to the speaker/participant if there it an email field for that speaker/participant.
BE-2810 TA: Switch Black Project color to Purple (for dark mode)
BE-2329 TA: Track LLM token use in Transcribe App Cloud
BE-2739 TA: Tracking used time on Transcribe App.
BE-2791 TA: Update look and feel of transcribe app

New or changed functionality in other platform components:

BE-2758 Add codec restrictions to the dial string for outbound
BE-2045 Add to AIVR API subReturn action.
BE-1782 Added Copilot Call Notes support.
BE-2777 Added info about LLM tokens to GET account new-billing method.
BE-2698 Added information regarding Member being Verified to the Call Notes
BE-2695 Added realTimeTranscriptText to AIVR Session and store the transcript from real-time transcription in it.
BE-1757 Added required features for Inbound Bot.
BE-2821 Added secondary voice connector id to AIVR App api.
BE-2768 Added speaker email value in the meeting data (ASR Meeting API)
BE-2404 Added support for language switching in telephony bot.
BE-1761 Added to AIVR API warmTransfer action.
BE-2767 Added to Edge Configuration parameters for connecting to Zoom Service.
QA-1319 Admin Tool: Added email validation on add account page.
BE-2657 Admin Tool: Convert API Use chart to chart-js
BE-2114 Check if we are within phone number quota on AWS before creating a new one
BE-2749 Check Voice Connector number against quota before creating a new one.
BE-2765 Demo: Collect events to Matomo from the voicebot part of the demo
BE-2761 For confGroups of type other than Transcription do not delete them if the creator is deleted
BE-2726 Get available information about participants of the specified Zoom meeting.
BE-2776 Grouping segments by meeting and attaching meeting metadata for /asr/meeting/llm/query
BE-2712 If the rex sessions are stuck (in error) do not count them towards the live session count during shutdown
BE-2759 If user deletion fails for any reason, we should still audit log it.
BE-2737 In AIVR, send vg::sub-return Event to Lua script (Freeswitch)
BE-2799 Introduced wew llmSettings.prompts for Meeting minutes.
BE-2733 Push data from Aircall Integration to the Copilot.
BE-2754 SA: Add extra confirmation before deleting a user who has the Admin role.
BE-2678 SA: Added API - GET /sa/offline/{id}/transcript
BE-2713 SA: Added context to the return values from GET /sa/call/agent
BE-2703 SA: Added contextId to the request for /sa/call
BE-2677 SA: Added download ability to the transcript view.
QA-1427 SA: Added search feature to transcript page.
BE-2689 SA: Addedd contextId to the GET /sa/call/{callId}
QA-1534 SA: New project gets default time settings from the Account.
BE-2774 SA: Supporting accounts with INVOICE billing on first login.
BE-2697 SA: When submitting /sa/offline request to as part of SA AIVR Integration set correct formatting parameters.
BE-2572 Send Caller Insight to Aircall as part of handling call.created webhook
BE-2790 Set the non-Speaker email value in the meeting data.
BE-2670 Transcribe App Client can get notified over websocket of any meeting join events instantaneously.
BE-2732 Using Pusher to push information to the Copilot Browser Extension.
BE-2736 Web Console: Added editable description to Edge Cluster
BE-2622 Web Console: AIVR Sessions table show which sessions have SA integration.
BE-2755 Web Console: Do not allow spaces in the names of AIVR Apps.
BE-2696 Web Console: Show real-time transcript text in the AIVR session detail view
BE-2747 Web Console: Showing AIVR App Id on the AIVR App detail drawer.
QA-1472 Web Console: Unable to add second URL in logic when its name is same as first.

Changes related to Integrity of Processing (fixes):

BE-2822 Fix - Can’t play audio from URL in bot logic callback
BE-2741 Fix - Duplicate data in meeting_vector database on Edge
BE-2727 Fix - Error when leaving a meeting (meeting join API).
BE-2721 Fix - Large vocabulary BUILT-IN grammar is not working in telephony bot
BE-2760 Fix - Origination data is not properly set for a new Voice Connector.
BE-2753 Fix - Outbound calling does not work anymore.
BE-2204 Fix - Percolator stop request ignored if sent immediately after start
BE-2806 Fix - The hidden checkbox stops working on ascalon TranscribeApp in some cases
BE-2817 Fix - Voice Connector quota check is not working OK
BE-2752 Fix - voiceConnectorId is not set on AIVR App if a phone gets added to AIVR App that did not have any phone
BE-2700 Fix - Weird Agent Composite values for the Demo Project /sa/calls
BE-2766 Fix prompt for Action Items.
BE-2729 Fix the shutdown of ascalon-asr-api
BE-2773 SA: Fix - Clicking on Terms of Service during SA Signup messes up signup
BE-2550 SA: Fix - Incorrect default time in file upload.
BE-2595 SA: Fix - JWT is not automatically inserted into the selected AIVR App.
QA-1475 SA: Fix - Pagination on Users page looks weird.
QA-1531 SA: Fix - Queue filter in Advanced search is not working for Uploaded file.
QA-1474 SA: Fix - Score filter is still available in advanced search even when score has been removed.
QA-1373 SA: Fix - The Continue button should remain disabled until all mandatory fields have been filled out.
QA-1541 SA: Fix - The login button should remain disabled until all required fields are filled.
QA-1437 SA: Fix - The role “Coach” is displayed as “QA” after being saved in Users.
QA-1480 SA: Fix - Unable to clear the description in General settings.
BE-2701 SA: Fix - undefined:undefined on the Topic chart
QA-1476 SA: Fix - User deletion is not working properly.
QA-1467 SA: Fix - User is able to set Persistence time more than 365 days in AIVR app.
QA-1446 SA: Fix - When a user goes back, they are redirected to the calls page instead of the Advanced Search.
BE-2699 SA: Fix - When I add a new keyword the dialog opens prepopulated with the name of the previous keyword I created
QA-1525 TA: Fix - Accepting only digits in the domain name for the “Allow signup with emails from the following domains:” field.
QA-1443 TA: Fix - Error message is wrong for the password reset page.
QA-1638 TA: Fix - Getting 500 server error when Users use Special characters in search bar on Advanced search page.
BE-1388 TA: Fix - Getting identical values of confidence when running transcript/{session-id} REAL-TIME
QA-1493 TA: Fix - HTML element showing in overview section on production.
QA-1613 TA: Fix - In Advance search Filter, tag is not working.
BE-2756 TA: Fix - List of transcripts on Home page loads twice
QA-1511 TA: Fix - LLM prompt box on transcript should not be available for shares.
QA-910 TA: Fix - Search by numbers is not working properly.
QA-1454 TA: Fix - Text “Undefined” populate when user left blank while signup
QA-1593 TA: Fix - The Advanced Search functionality breaking on searching more than one word.
QA-1461 TA: Fix - There should be a limit for the expiry time limit for shared transcript.
QA-1481 TA: Fix - There should be warning Dialog when user click on Back Button or Close Browser
BE-2330 TA: Fix - Transcribe is showing one-line per word.
QA-1636 TA: Fix - Unable to check the creator on the advanced search filter.
QA-1462 TA: Fix - Walk through wizard breaks when user switch language between walkthroughs.
QA-764 TA: Fix - When a whole sentence entered in search box, it is not working and overlapping with close icon.
BE-2828 Web Console: Fix - AIVR App creation allows you to pick phone numbers that are not yet ready for use.
QA-1425 Web Console: Fix - Double error showing when phone number purchase.
BE-2716 Web Console: Fix - Error page when trying to open AIVR session details.
QA-1513 Web Console: Fix - Transcript play time exceeds than actual time when playback speed increases.
QA-1575 Web Console: Fix - User is unable to Logout properly.
QA-1483 Web Console: Fix - User Management Search - When user searches with leading or trailing whitespace with email then no results come.
QA-1526 Web Console: Fix - When user switch context on Profile page then 2FA pop-up keeps showing each time context is changed.

Release 1.107.0

Key changes related to the core APIs

Changes in AIVR API to support Aircall integration
Changes in AIVR API to support multiple languages
Agents and Queues now correctly handles in SA APIs and SA App
Improved Advanced Search in SA App
Ability to use single session on websocket receiving server for both left and right transcription channels.
Ability to add external UUID to any transcription session
Full HA setups for FreeSWITCH and Kamailio

Key changes related to Transcribe APP

Meeting metadata used in LLM Playground
LLM Query uses history of previous Q and As
More detail for the relevant meeting display in LLM Query
Enhancements to text file download

New or changed functionality in the Transcribe App:

BE-2633 TA: Add metadata information to LLM Playground query
BE-2724 TA: Add option to include header in the downloaded text file
BE-2651 TA: Add Project information to the relevant-meeting list in LLM Query page
BE-2687 TA: API Key entry boxes should be set to not remember history
BE-2634 TA: For text transcript download, add configurable transcript timestamp
BE-2725 TA: Improve the rules to show the speakers and non-speaking participants
BE-2577 TA: Indicate the relevance of the relevantMeetings returned from /asr/meeting/llm/query
BE-2717 TA: Make sure that the transcript submitted to LLM Playground has timestamps (interval parameter)
BE-2601 TA: Remember sorting on home page when returning to it from open meeting details
BE-2589 TA: Replace HighCharts WordCloud with another library
BE-2570 TA: Show the scope of the LLM query in the Answer

New or changed functionality in other platform components:

BE-2673 Web Console Add search by email to user management page
BE-2617 Web Console: Add 3 new Aircall parameters to the AIVR App settings
BE-2605 Web Console: Add language to the logic settings on the AIVR App
BE-2604 Web Console: In AIVR App settings - support multiple voices (one per language)
BE-2598 Web Console: Make number of rows per page in the Phone Apps table configurable
BE-2667 Web Console: New Voice Connector status column in AIVR App view
BE-2666 Web Console: Shortcuts from Phone Management page plus modified status field
BE-2669 Web Console: Show event metadata (if present) for AIVR Session Events
BE-2012 Web Console: Transcribe+ now uses /sa/offline API
BE-2649 Web Console: When querying AIVR (telephone) session set the limit parameter to 500
BE-2675 SA: Add a refresh button to the page with the list of Calls
BE-2635 SA: Add Description to Project
BE-2631 SA: Add tooltips on 1D, 7D, 30D
BE-2607 SA: Improvements to AIVR App selector
BE-2608 SA: Make the selected Project indicator have 2 instead of 1 letter
BE-2380 SA: On project general settings highlight the color and show time settings preview
BE-2606 SA: Preselect time settings on the page where we create new project
BE-2621 SA: Show headline in /sa/call search results if the search included text search
BE-2636 SA: Show project name on the Integration settings page
BE-2590 SA: Show the integration Icon on the Project selection list
BE-2647 SA: Use the new GET /sa/call/agent to map agent_user_id to the name of the Agent
BE-2646 SA: Use the new GET /sa/call/queue method to provide names for queues in the Advanced Search Queue filter
BE-2645 SA: When querying fields for advanced search pass the contextId for the current Project
BE-2611 Admin Tool: Convert Account Activity chart to Chart.js
BE-2664 Add a verify query parameter to GET/account/{uuid}/phone-number
BE-2643 Add contexId parameter to GET /sa/call/search/fields
BE-2650 Add contextId to the content of the relevantMeetings returned from GET /asr/meeting/llm/query
BE-2603 Add default TTS/ASR language for logic in AIVR App
BE-2623 Add externalSaSession flag to AIVR Session
BE-2648 Add limit parameter to the API that returns AIVR sessions
BE-2660 Add metadata field to the AIVR Event
BE-2665 Add new verify query parameter to GET /aivr-app
BE-2720 Add optional metadata parameter to method that GETs meeting transcript
BE-2614 Add to AIVR App: aircallApiUsername and aircallApiPassword
BE-1783 AIVR API now can enable real-time transcription on transferred leg
BE-2221 Allow the AIVR logic to send an alternative callback url in the response
BE-2503 Associate and external ID to transcribe session on post in order to be able to retrieve the session using that ID later
BE-2618 DAO for a database that will be storing data about recordings from S3 that need to be transcribed
BE-2602 Define default voice for each language supported by AIVR App
BE-2559 Delete embedding vectors if meeting is deleted.
BE-2296 Deploy HA Kamailio Setup in Cloud
BE-2644 Implement GET /sa/call/agent
BE-2642 Implement GET /sa/call/queue
BE-2571 In FreeSwitch transcribe proxy remove REDIRECT_URI and create SIP uri used in deflect from SIP URI name and DESTINATION DOMAIN
BE-2715 More accurate control over the number of tokens submitted to LLM Service with all the embeddings
BE-2563 Move meeting-join docker image from public repo to private repo
BE-2526 New settings for voiceconnector origination and termination
BE-2482 Pass intents from AIVR Session to /sa/offline via topics in /sa/call
BE-1366 RedisConnectionException: Unable to init enough connections amount! Only 0 of 24 were initialized.
BE-2627 Remove xml:lang from the SSML template in audio-server
BE-2641 Report Aircall Agent to the /sa/call
BE-2640 Report Aircall Queue to the /sa/call
BE-2620 Return headline as part of the /sa/call search results
BE-2573 Send Call Notes to Aircall when AIVR call ends
BE-2572 Send Caller Insight to Aircall as part of handling call.created webhook
BE-2531 Send multiple asr session results to a single websocket receiver server session
BE-2659 Split post-processing of AIVR sessions into immediate step and the delayed step
BE-1786 Store copilot notes in /sa/call
BE-2579 Store meeting speaker timeline in the database
BE-2676 Support follow up queries in the /asr/meeting/llm/query - add history parameter and use it
BE-2406 Support language switching setting in telephony bot logic callback
BE-2652 Sync AIVR session with Aircall callId at transfer
BE-2583 Use meeting.join.puppeteer.imagePullSecret property to pull puppeteer docker image

Changes related to Integrity of Processing (fixes):

BE-2034 Admin Tool: Fix - failed to decode request body: organization name “telegraf” not found”
BE-2600 Fix - Aircall webhook is unable to find an AIVR session
BE-2674 Fix - Error when saving queue field data in /sa/call API
BE-2705 Fix - fail to get fsPodName from the db for outbound calls
BE-2628 Fix - In case of AIVR/SA integration the saConfig is retrieved from AIVR App context instead of from the SA Project
BE-1354 Fix - NatsEndpoint fails to reconnect to NATS when its connection is lost
BE-2637 Fix - realTimeAsrTranscribeSession is set on AIVR Session, but processing after the call terminates claims that “no RT asr transcribe session”
BE-2629 Fix - Still seeing default text in the sa_call table
BE-2638 Fix - Transcript text is not being saved into text field in sa_call
BE-2592 Fix filebeat logging on cloud
BE-2594 SA: Fix - Incorrect message shown after Voicebot Integration Project created
BE-2373 SA: Fix - Name validations for project settings are required.
BE-2691 SA: Fix - Swapped channels in SA AIVR Integration
BE-2560 TA: Fix - After latest change the Meeting Bot does not distinguish between progress Processing or Done
BE-2630 TA: Fix - Meeting search seems to not always return the newest first
QA-1485 TA: Fix - Error page when opening rediarized transcript (partial fix)
BE-2692 Web Console: Fix - Error when listing some sessions Call Session view
BE-2693 Web Console: Fix - Filter on the App Name in the Call Sessions table does not work
BE-2525 Web Console: Fix - Label validation issues in Transcribe+
BE-2534 Web Console: Fix - Provide correct value of the SIP URI on QA

Release 1.106.0

Key changes related to the core APIs

Offline model (omega) trained on additional data from call center calls (health insurance, retail customer support)
Improved diarization
Improvements in /asr/meeting/join API (for Meeting Bot)
Improvements to /asr/meeting/llm/query API (return relevant meetings)
Outbound dialing needs to be enabled per account (it is a compliance feature)
Added AIVR (Voicebot) Integration to Speech Analytics App
Added webhook to Aircall integration
Added HA Freeswitch support
Collecting TTS character and LLM token usage
Fixed a configuration issue where Grafana was not available on Edge

Key changes related to Transcribe APP

Added Meeting Bot that can join a meeting and record/transcribe it (beta)
Added Rediarize option
LLM Playground and LLM Query available on Cloud
- Note there is a limit of 16K tokens - about 1.5h of audio
LLM Query results now show which transcripts were used in answering the query
Added LLM Service authorization header settings
Improved Action Items LLM query
Show which users have Zoom Meeting Assistant installed

New or changed functionality in the Transcribe App:

BE-2468 TA: Add a meeting_bot tag to meetings recorded using Bot
BE-2470 TA: Add BOT selector on the home page
BE-2345 TA: Add Meeting Bot
BE-2412 TA: Add Re-diarize option for a meeting
QA-766 TA: Added search to time-zone selector
QA-1281 TA: Better error message when inviting user with invalid domain
BE-2545 TA: Do not hide LLM Playground on the Cloud
BE-2546 TA: Do not hide LLM Query on the Cloud
QA-1377 TA: Fix after-login redirect URL on Edge
BE-2508 TA: For first time login, after creating first project, we should take user to the page with the Zoom Meeting Assistant
BE-2529 TA: Handle 429 response from /llm/chat API and /asr/meeting/llm/query
BE-2467 TA: Identify Meeting Bot recordings as such in Transcribe App
BE-1998 TA: Improved 404 Error handling
BE-2394 TA: Improvements to LLM Playground UI
BE-2495 TA: In LLM settings add Authorization header settings
BE-1094 TA: Information about the use of Shared links is displayed
BE-2514 TA: Make the request for Action Items be similar to the request made from the LLM Playground
BE-2487 TA: Modify the LLM Query page to have the same look and feel as the new LLM Playground
BE-2392 TA: New widgets to set number of speakers for Saving microphone recording
BE-2391 TA: New widgets to set number of speakers for Upload
QA-1416 TA: Remove language selector on Project Settings page on Edge
BE-2339 TA: Show on Users page if the user has Zoom Meeting Assistant installed
BE-2486 TA: Show relevant meetings if they are returned by the /llm/query API
BE-2207 TA: Show share usage
QA-321 TA: When a new user creates his first project there is now an option for change language for page translation

New or changed functionality in other platform components:

BE-2524 Add a failsafe to all meeting bots
BE-2481 Add aircallId field to AIVR Session
BE-2480 Add aircallWebhookToken to AIVR App
BE-2367 Add audio websocket URL to response from POST aivr logic callback
BE-2494 Add authHeader parameter to llmSettings in /cluster API methods
BE-2343 Add avirAppId and aivrSessionId to /sa/call API
BE-2469 Add BOT value to AUDIO_SRC in meeting search API
BE-2520 Add date field to relevantMeetings
BE-2383 Add diarization parameter to re-run meeting API
BE-2477 Add fsPodName to the POST /aivr
BE-2368 Add read-only field llmCopilotNotesPrompt to saConfig
BE-2444 Add relevantMeetings to the response from GET /asr/meeting/llm/query
BE-2439 Add reviewNotes field to /sa/call API
BE-2351 Add to GET /sa/call API ability to query by AIVR App Ids
BE-2400 Add to storage measurement an explicit multiply column
BE-2452 Allow Outbound Dialing only of OutboundDialing is set on the account
BE-1785 At end of call, AIVR to submit real-time transcript to GPT to generate copilot notes
BE-2232 Better handling of ServiceUnavailableException from cloud Influxdb
BE-2338 Collect LLM statistics in InfluxDB
BE-2042 Collect TTS statistics in InfluxDB
BE-2516 Do not launch new /meeting/join if there are already N running meeting join K8s tasks
BE-2466 Enforce Meeting Join limits on Transcribe App Cloud accounts
BE-2458 For all AIVR sessions use POST /sa/offline/call/{callId}
BE-2437 Generate call summary using llmSummaryPrompt setting in saConfig (offline)
BE-2536 If the underlying LLM service behind /asr/meeting/llm/query API returns 429 error, the API should also return 429
BE-2479 Implement POST /public/awebhook/aircall
BE-2362 Implemented PUT /asr/meeting/{meetingId}/leave
BE-2270 Improved AIVR Session Start Time
BE-2358 Increase diarization.maxSpeakers from 12 to 20
BE-2548 Increase max number audio on meeting API from 25 to 35
BE-2438 Increase SA shortSummary length to 2048
BE-2533 Make cleanup property configurable
BE-2390 Modifications to how we configure Voice Connectors in Telco Service
BE-2352 Modify GET /account/{uuid}/new-billing to obtain usage data from influx DG rather than from the account
BE-2411 Modify pagination style for GET /sa/call
BE-2433 New debug event and tweaks to the existing events in meeting join API
BE-2482 Pass intents from aIVR Session to /sa/offline via topics in /sa/call
BE-2549 POST /llm/chat now works in Cloud as well as Edge
BE-2213 Removed Prompt Manager from audio server API
BE-2501 Request to /llm/chat should use the authHeader settings from llmSettings
BE-2502 Request to compute action items and any other llm requests for the transcript should use authHeader from llmSettings
BE-2500 Request to Embeddings API should use the new authHeader parameter
BE-2463 Return different 404 error code in case “No pods found for job: meeting-join-krxuobktp3kyez2uuqo3”
BE-2371 SA: Add filtering by Agents
BE-2422 SA: Add filtering on Call List page for Agent, Queue, and Resolution
BE-2374 SA: Added a placeholder image for default profile avatar.
BE-2228 SA: Added AIVR (Voicebot) Integration
QA-1289 SA: Added success/alert notification after settings are updated or saved.
BE-2370 SA: After selecting a custom date, show it on the page
QA-1316 SA: Highlight Project name while deleting project.
BE-2423 SA: Improve the Date Picker (e.g. on the Call List Page)
BE-2380 SA: On project general settings highlight the color and show time settings preview
BE-2420 SA: Remove the search box from the Call list page
BE-2455 SA: Show notes about the Call
BE-2378 SA: Show preview of Time Settings
BE-2507 Set aivrSessionId and AivrAppId on /sa/call when we create it at the end of the aivr session even if there is no AIVR Integration
QA-1390 Set the correct text on sa/call at the end of /sa/offline processing
BE-2377 Support for multiple FreeSWITCH services
BE-2506 Switch to new sentence embedding model in ml-svc
BE-1557 Track use of /asr/meeting API for billing
BE-2526 Two new settings for voiceconnector origination and termination
BE-2519 Update the omega model to the latest version
BE-2543 Use the weighted average on the similarity scores on agent detection
BE-2440 Voicebot logic to start real-time transcription sessions and submit audio WS URL and asr session IDs in response
BE-2429 Voicebot: Detect intents using examples
BE-2409 Voicebot: Translate prompt on the fly and store to firestore
BE-2344 Web Console: Add Calls page to Telephony Bot mode
BE-2297 Web Console: Add filter by name to the list of Edge Configurations
BE-2307 Web Console: Add view like ‘kubectl get pods’
QA-1042 Web Console: Added an option for Log out from all devices under account settings.
BE-2454 Web Console: In the list of phone Apps mark those that have SA Project integration

Changes related to Integrity of Processing (fixes):

BE-2478 Address issue where some users end up in Main Org. in Grafana
BE-2385 Admin Tool: Fix - CMP user get 401 error when retrieving user data from some other account
BE-2427 Fix - Cannot store some data in sa_call table
BE-2456 Fix - GET /asr/transcribe/{sid}/transcript?format=text” hangs is many hints are used and audio is over 30 minutes
BE-2498 Fix - Meeting join is not working on Edge
BE-2204 Fix - Percolator stop request ignored if sent immediately after start
BE-2415 Fix - Sentiment values have now narrower range than they used to be
QA-1388 Fix Grafana on Edge
BE-2457 Fix: Bot cannot join meeting in multiple participant scenario
BE-2280 Remove speakers field from POST /sa/offline/call/{callId} method
BE-2474 SA: Fix duration filter
QA-1206 SA: Fix - Advanced Search is not working
BE-2381 SA: Fix - Agent statistics page needs a spinner as soon as one of the time periods get selected
BE-2348 SA: Fix - Call duration has wrong values
BE-2418 SA: Fix - Cancel on Edit use simply closes the window without restoring the state
BE-2379 SA: Fix - Configuring of Silence and Overtalk thresholds is messed up
BE-2419 SA: Fix - Double loading of the call details page
QA-1318 SA: Fix - Duration filter is not working in Advanced search.
BE-2416 SA: Fix - Icons and Letters are getting cut off in the Call List view
BE-2373 SA: Fix - Name validations for project settings are required.
BE-2346 SA: Fix - Next button works only on the 1st call loaded
QA-1309 SA: Fix - Pagination on Agents page have Chinese letters instead of “per page”.
BE-2369 SA: Fix - Style issue on number of pages selection in modal.
QA-1275 SA: Fix - Unknown symbols instead of “Next” and “Back” buttons
QA-1066 SA: Fix - User is unable to change profile photo
BE-2372 SA: Fix - When using current time as default time, if I don’t change the value it gives me the error: ‘Start Time is required’, it should be set as current date by default.
QA-1298 TA: Fix - Admin User is unable to update the LLM setting.
QA-1277 TA: Fix - After submitting a microphone recording, no success or upload message is shown. It should display a success message.
BE-2331 TA: Fix - Browser Capture Pop-Up seems to load bunch of stuff that it does not need
QA-1304 TA: Fix - For browser share and microphone capture recording, the Success message is not displaying after saving the recording.
QA-1357 TA: Fix - LLM Query is not working
QA-1359 TA: Fix - Redux is not clearing up in case of session logout
QA-1196 TA: Fix - Sometimes, the login process gets stuck on the login page.
QA-1384 TA: Fix - Three dot menu are not visible for the account users table.
QA-965 TA: Fix - Within account share is not working properly.
QA-1286 Web Console: Fix - “Create” button for Context should be disabled after first click and loader should show
QA-1280 Web Console: Fix - Edit issues on API Secrets page
BE-2322 Web Console: Fix - Error when I modify AIVR APP (Change CSID Callback from Path to Query)
QA-1282 Web Console: Fix - Getting error message when user starts recoding but does not save it
QA-1181 Web Console: Fix - GREG: When users try to filter True column of Interpretation and Review Status getting blank page
QA-1358 Web Console: Fix - Pagination Control/Sessions per page is not working under Account management.
QA-1178 Web Console: Fix - Text “Undefined” coming in “Main DNIS” dropdown
QA-1180 Web Console: Fix - Unable to Edit AIVR App Settings getting 500 error
BE-2461 Web Console: Fix - Unable to log into grafana on Edge on dev and QA
QA-1272 Web Console: Fix - When user logout and login again, get Authorization error and login failed

Release 1.105.0

Key changes related to the core APIs

Offline model (omega) trained on additional data from call center calls (health insurance, retail customer support)
Telephony Bot API now supports outbound calling and setting ANI caller id on transfer
Added a parameter that controls language detection in offline transcribe API
Added ability to send real-time transcription results to a websocket server (previously only supported websocket client)
Web Console now shows all login sessions and supports forced logout.

Key changes related to Transcribe APP

Added basic LLM Query over data stored in vector database
User-level time settings (previously time settings were only account-wide)
Korean transcription should now be less likely to shift to translation
Summary LLM prompts are customizable on Edge

New or changed functionality in the Transcribe App:

QA-1230 TA: Active login sessions now displayed and highlighted at the top by default.
BE-2276 TA: Add 3 more LLM prompts to LLM settings
BE-2224 TA: Add ‘copy to clipboard’ for action items
BE-2278 TA: Add time settings also on User Profile
QA-1237 TA: Added an option for Logout from all devices on My Login sessions.
BE-2217 TA: Added basic Vector LLM Query
BE-2284 TA: Apply MD formatting to the section summaries
BE-2314 TA: If a user clicks the LLM Query button on the home page - take them to the LLM settings if needed
BE-2240 TA: Implement page for Vector based querying - Edge Only
BE-2196 TA: Make sure that whenever we discard the login session in the browser we do call logout API
BE-2218 TA: Provide immediate feedback of the Start mic capture button being clicked
BE-2282 TA: Use the customer overrides (if specified) for the summary prompts

New or changed functionality in other platform components:

BE-2198 Web Console: Add login sessions view in Account Profile page
BE-2197 Web Console: Make sure that whenever we discard the login session in the browser we do call logout API
BE-2001 Web Console: Notify about inability to to microphone capture on HTTP urls
QA-1274 Web Console: Removed Prompt and TTS Settings
BE-2233 Web Console: Show the list of languages sorted
BE-2229 Web Console: Use the new change password API
BE-2199 SA: Added created date and UUID to the page with Project details
BE-2200 SA: Added name of the project to the Delete dialog
BE-2256 SA: Contact Support now opens in a new browser tab
BE-2290 SA: Implement AIVR Integration (configuration)
BE-2206 SA: Improve PII Redaction Config page
BE-2255 SA: Use the Account Time settings as the time defaults when creating a new Project
BE-2254 SA: Use the new change password API in SA App
BE-2252 SA: Hide (comment out) the setting for Age detection
BE-2299 Demo: Add a note about the models used in the Voicebot Demo
BE-2300 Demo: Copy to clipboard icons need tool-tip and there needs to be confirmation after copy
BE-2211 Admin Tool: Add tab/page to control SA activation
BE-2301 Add languageDetection parameter to asr settings in the transcribe API
BE-1749 Add Outbound Calling to AIVR API
BE-2277 Add to cluster API new LLM Settings for the Embeddings
BE-2177 Added support for 11labs voices in TTS
BE-2225 Added transfer to extension option in the AIVR transfer action
BE-2291 AIVR Transfer should give option to use DNIs or Original ANI for the caller id for the transferred call
BE-2312 Better handling of HTTP 502 from Fusebill
BE-2279 Changes on agent field in request to POST /sa/call
BE-2085 Do not return “user not found” from the forgot password API
BE-1972 Hide/Show features of the Speech Analytics App based on the settings on the account
BE-2239 Implement POST /asr/meeting/llm/query
BE-2286 Improve (speedup) lookup of AIVR App by DNIS
BE-2216 Improve diarization sentence boundary
BE-2209 Language detection for offline sessions using Omega model
BE-2273 LLM meeting section summary now returned raw
BE-2193 Modified GET /greg/experiment to return only the experiments that are from the specified Context
BE-673 Modify password change API to encrypt password parameters
BE-2081 Move caching of TTS prompts from memory to redis
BE-2185 Move mod_xml_curl service to fssk
BE-1791 Omega model trained on Health Insurance Call Center data
BE-2230 Speedup the response time of POST /asr/transcribe/async
BE-2191 Support ‘external’ websocket mode for /asr/transcribe/async and /asr/recognize/async
BE-2192 Support new returned Content type - metadata
BE-2261 Tie AIVR App and Context to form AIVR Integration (Context side)
BE-1680 Update jdk version to 17 in docker images

Changes related to Integrity of Processing (fixes):

QA-1185 TA Edge: Fix - Unable to recognize transcript’s project on homepage.
QA-1252 TA Edge: Fix - Port Number is missing in invitation link
BE-2269 TA: Fix - Action Items are no longer included in the PDF and Docx
QA-1198 TA: Fix - Advanced Search View does not update after clicking ReRun.
QA-1263 TA: Fix - After deleting any share, it still shows in the table until the user performs a hard refresh of the page
QA-1240 TA: Fix - Downloaded pdf for Korean transcript does not open in Acrobat Reader
QA-1248 TA: Fix - Duplicate transcripts are showing in advance search.
BE-2275 TA: Fix - Korean sometimes being translated in addition to transcribed
BE-2274 TA: Fix - LLM Playground needs to be able to get transcript even if user is User role and has no Download permission
QA-1279 TA: Fix - Microphone recording and browser share menu should not be shown under actions for languages that do not have permissions for browser capture and microphone recording.
QA-1126 TA: Fix - Null projects displayed for some meetings.
QA-1199 TA: Fix - Some transcripts are missing a options under the 3-dot menu.
QA-1232 Ta: Fix - Sorting is not working on MY Login sessions page.
QA-1219 TA: Fix - The ‘Incorrect Password’ message should appear when a user tries to download the password recovery key with the wrong password.
BE-838 TA: Fix - Tracking of time used does not seem to work anymore
QA-1242 TA: Fix - User gets stuck on Upload page after submitting a file.
QA-912 TA: Fix - User is able to upload the audio file even when the allotted storage is fully used.
BE-2313 TA: Fix - User unable to move recording from one project to another
QA-1205 TA: Fix - Warning not showing for upload of large file.
BE-2169 Web Console: Fix - Clicking on the date range in Logs page triggers an error in the console
QA-1220 Web Console: Fix - New user Account getting error on log search
BE-2271 Web Console: Fix - Output AIVR prompts are not expanding
BE-2219 Web Console: Fix - recorded audio chart should start from 0:00
QA-1243 Web Console: Fix - Unable to edit ‘App Name’
QA-1273 Web Console: Fix - When user click on the No. of node of ‘No edge deployments loaded’ table getting blank page
QA-1258 SA: Fix - Audio forward/backward buttons are not working as expected.
BE-2316 SA: Fix - Avatar not getting saved
BE-2250 SA: Fix - Back Button should take us back to Settings overview and not to main page
QA-1226 SA: Fix - Fails to show the new project wizard if there are not Projects on account
BE-2260 SA: Fix - Missing parameters for uploaded Call
QA-1159 SA: Fix - User is unable to select the time and time dropdown text is overlap.
QA-1068 SA: Fix - User unable to update location , Company Position, and Name, as save button disabled.
QA-1133 SA: Fix - Weird behavior on the configurations page.
BE-2288 Fix - AIVR stopped listening in the middle of a my answer
BE-2208 Fix - asr-api fails to submit DEL callback request after disconnect is processed
BE-2181 Fix - ml-svc performance degradation since release 103 after we upgrade to python 3.11
BE-2214 Demo: Fix - Getting 401 error when trying to send a link from the demo app
QA-1213 Admin Tool: Fix - Showing a blank page when the admin user tries to check the API usage.

Release 1.104.0

Key changes related to the core APIs

Add diarization to Whisper model in offline task
Autoscale real-time REX on GCP
Automatically detect agent channel in offline SA API
Improved Kappa (real-time) model trained on additional call-center data (health insurance domain)
Switch Analytics App from /sa API to /sa/offline APIs
Extend the use of allowSignupsFromDomain to allowed user invites

Key changes related to Transcribe APP

Add Korean Language option for Edge
Enhanced LLM Playground to support temperature setting and markdown rendering
Added “Automatic” setting for number of speakers in diarization
In Advanced Search added Relative Time filter
Add pages with list of login sessions (for user and for all account)

New or changed functionality in the Transcribe App:

BE-2133 TA: Add “Go to Home” button on “Something Went Wrong” page
BE-2202 TA: Add a 3 second timeout on the request to ipfy
BE-2134 TA: Add Korean Language option for Edge
BE-2078 TA: Add page with list of all logged-in sessions on the account
BE-2182 TA: Add support for m4a files in file upload
BE-2155 TA: Add temperature to LLM Settings
QA-680 TA: Better reporting of Recompute in progress on Advanced Search page
BE-2125 TA: LLM playground able to render markdown format in LLM responses
BE-2190 TA: Modify message template used LLM Playground to work with Mixtral
BE-2141 TA: Modify the number of speakers selector for diarization on Upload
BE-2184 TA: Remove diarization from Live Microphone recording preview
BE-2118 TA: Remove previous error message from the login screen as soon as the user clicks the login button
QA-848 TA: Show all login sessions of the current user
BE-2136 TA: Show PC Name when hovering over Installed-Connected
BE-2164 TA: Support Relative Time search filter on Start Time in Advanced Search
BE-1885 TA: Use Mixtral for Summarization on Edge

New or changed functionality in other platform components:

BE-2142 Add ability to save Meeting Search queries
BE-1698 Add diarization to Whisper model in offline task
BE-2062 Add percolatorThreshold to the output response in AIVR callback
BE-2139 Add regex syntax and group references to the url prefix matching in grammar fetcher
BE-2137 Add RelTimeTerm query term to Advanced Meeting Search API
BE-2138 Add RelTimeTerm query term to Call Search API
BE-2144 Add temperature parameter to llmSettings in the /cluster API
BE-2092 Admin Tool: Show login error messages to user
BE-1601 API to email link to the results of Voicebot Demo session - make the link url configurable
BE-2127 Autoscale real-time REX on GCP
BE-2120 Demo Voicebot: Show the page with trancript/audio/sidepanel immediately after the end of the call
BE-2158 Demo Voicebot: Support Spanish version of Casey demo
BE-2187 Demo: Use the new urlLink parameter in the API used to send the link
BE-2089 Detect agent channel in offline SA API
BE-2175 Extend the use of allowSignupsFromDomain to user invites
BE-2135 Improve /llm/chat API to handle transcript longer than the token limit
BE-1792 Improved Kappa (real-time) model trained on additional call-center data (health insurance domain)
BE-2041 Limit maximum number of nonces that can be active for a login session at any time
BE-1678 Migrate Java Applications from Java 11 to Java 17
BE-2172 Modify the GET method for the Business Config API which takes Auth Token
BE-2108 Monitor and restart asr container if it loses GPU
BE-2126 Retire old billing-utility
BE-2117 Return a specific http error response if we get ResourceLimitExceededException when creating new Voice Connector.
BE-2149 SA: Add filter on User Role
BE-2129 SA: Add JWT token generation to SA
BE-2157 SA: Add option to Delete a user
BE-2156 SA: Add option to Edit the User
BE-2057 SA: Invoke correct new APIs when uploading single file on SA App
BE-1712 SA: New Advanced Search Page
BE-1794 Switch Analytics App from /sa API to /sa/offline APIs
BE-1997 Tie collecting storage information to activity on account
BE-2109 Upgrade openai package to version 1.*
QA-1165 Web Console: Add a filter for the 2FA enabled on the user management page.
QA-1188 Web Console: Add Calendar to date selection in Business Config
BE-2080 Web Console: Add sorting on columns showing available Edge configurations
BE-2121 Web Console: Edge deployment filter - bring the search box in focus as soon as I open the filter
BE-2075 Web Console: Show telcoData.fsUUID on the AIVR session details page
BE-2176 Web Console: Show value of allowSignupsFromDomain in account profile

Changes related to Integrity of Processing (fixes):

BE-98 Fix - Bug in telco service with Apps that are sip-only and we try add a phone number
BE-2170 Fix - Float overflow issue in rex RNNT search
BE-2166 Fix - Incorrect column count: expected 1, actual 38 in getCallCount in /sa/call API
BE-1971 Fix - NoSuchFieldError: LUCENE_7_7_3
BE-2167 Fix - Speakers merged into streaks of other speaker utterances in /sa/offline API
QA-1133 SA: Fix - Weird behavior on the configurations page.
QA-1200 TA Edge: Fix - Invitation email is not coming to the provided email address.
QA-1236 TA: Fix - After editing any expired shared transcript to Never Expire, the save button is not working.
QA-1223 TA: Fix - Back button is not working on Walkthrough Wizard
QA-1153 TA: Fix - Blank page is showing for the login for one account.
QA-1211 TA: Fix - Hover msg over Never share Expiry status is wrong.
BE-1989 TA: Fix - Inconsistent Project membership as shown in the UI
QA-1076 TA: Fix - Personal projects are not showing in the destination projects while moving any transcript.
QA-1122 TA: Fix - Resend invites to bad email address throwing the something went wrong issue.
BE-2146 TA: Fix - Right-click not working on Edge
QA-1227 TA: Fix - Showing the American flag for all project language projects.
QA-1197 TA: Fix - Specific No. of Speakers field is accepting blank inputs.
BE-1987 TA: Fix - Stuck on Import/Upload page after Submit
QA-1231 TA: Fix - There is no limit given for the Time range in advanced search.
QA-1209 TA: Fix - Walkthrough wizards not working properly on 90% page zoom.
QA-586 TA: Fix - When a user clicks on “edit” of a Share it automatically increases its “Expires in” time by a day.
QA-1174 Web Console: Fix - On Experiment Browser section- ‘Apply filter’ button is not working, user getting ‘e.preventDefault’ is not a function.
QA-1191 Web Console: Fix - Only one success message should be displayed when adding a ‘Business Configuration’.
QA-1186 Web Console: Fix - Past date entry should not allowed in adding Business Config records.
BE-2148 Web Console: Fix - Unable to save the URL of the outbound AIVR logic
QA-1176 Web Console: Fix - User getting blank page, either clicked on True text or ascending- descending filter or when searching something.
QA-1177 Web Console: Fix - When a user click on copy icon under A of Utterance, he does not get any copied data but getting Type Error
QA-1179 Web Console: Fix - When users try to delete ID’s under Interpretation column by clicking on delete icon, it’ s not getting deleted.
QA-1175 Web Console: Fix - When users try to edit any ID under True column, since at the time of editing Reset and Save button is disable but still it is clickable.

Release 1.103.1

New or changed functionality in the Transcribe App:

BE-2143: (TA) Show Recompute option also if status Error under certain conditions

Changes related to Integrity of Processing (fixes):

BE-2163: Fix - Incorrect nginx routing for auth-svc from admin app

Release 1.103.0

Key changes related to the core APIs

Support for url rewrite and HTTP proxy for grammar fetching for /asr/recognize API and for MRCP ASR.
Switch Speech Analytics App from old /sa api to new /sa/offline API
New APIs to manage user login sessions
New API to define business specific settings for a Voicebot

Key changes related to Transcribe APP

LLM Playground for Edge Deployments - ask a LLM any questions about the transcript.
Fix for negative duration of words which was impacting recomputing data from transcripts.

New or changed functionality in the Transcribe App:

BE-2068 TA: Add LLM Playground page to Transcribe App (on Edge only)
QA-1054 TA: Add validator for keyword group names
BE-2060 TA: Option to set share within account to not expire

New or changed functionality in other platform components:

BE-2101 /sa/offline API - generate talk and overtalk data for each speaker
BE-2076 Add AIVR API method to reserve rex for already started outbound aivr session
BE-1953 Add AIVR Business Config API with opening-hours information and other info
BE-2036 Add aivrAppId to the initial callback for the AIVR session (also the websocket version)
QA-1071 Add API to logout user from some/all login sessions
QA-1070 Add API to show User’s all login sessions
BE-2061 Add audioOffset option to audio in /sa/offline
BE-1878 Add localization in llm-svc
BE-2044 Add parent AIVR session to POST /aivr/dial/{destination}
BE-1917 Add to Account a read-only field to store customization settings for SA App
BE-1974 Add to Account API saAppHiddenFeatures
BE-2020 Add version field to /sa/call
BE-2033 Associate Business Config with AIVR App
BE-1850 Casey Demo Voicebot in Spanish
BE-2058 Configure demoPhoneNumber for demo app on dev qa and prod environments
BE-2011 Convert auth-svc from tomcat+war to SpringBoot
BE-1985 Enhancements to grammar fetcher (http proxy, url rewrite)
BE-1972 Hide/Show features of the Speech Analytics App based on the settings on the account
BE-1909 Hook up Speech Analytics App to Sentry
BE-1954 Implement API to query existing login sessions
BE-2035 Implement GET /aivr-app/{aivrAppId}/business
BE-1992 Implement GET and DELETE /auth/login/{sessionId}
BE-1497 Implement getContextStats method on SearchableMeetingDao that returns statistics for contexts
BE-2074 Implement POST /llm/chat API
BE-1598 Implement proper queue priority for transcription
BE-2003 Implement safe shutdown for freeswitch/fssk pod
BE-2031 Improve nonce handling in Customer Management Portal (Admin Tool)
BE-2030 Improve nonce handling in Customer Portal
BE-2032 Improve nonce handling in Speech Analytics App
BE-2116 In Share APIs – make expires value of 0 mean that the share never expires
BE-2052 Modify REX to stop storing usages in Redis
BE-2051 Move sentryEnvironment to global setting in onprem-cluster-deployment task
BE-1073 New API: POST /asr/meeting/{meetingId}/rerun
BE-1520 Stop storing usages in Redis after billing-utility is no longer needed on production
BE-1981 Support cluster specific service account key in onprem-cluster-deployment task
BE-1794 Switch Analytics App from /sa API to /sa/offline APIs
BE-1966 Update k8sNodes in firestore automatically when changing cluster setup
BE-2069 Web Console: add “outbound” option to pull-down for logic
BE-1994 Web Console: Add audio download button to AIVR call audio page
BE-2038 Web Console: Add page to configure Business settings for use in AIVR
BE-2006 Web Console: include ;transport=tcp in the sip URI shown in Telephony Bot App
BE-2002 Web Console: Place cursor and focus on the highlighted textbox right after the context dialog is shown
BE-1492 Web Console: Show Edge configuration in the table that shows all Edge deployments
BE-1951 Web Console: Telephony Bot mode in Call Session table - make it remember previous state when returning from the detailed session view

Changes related to Integrity of Processing (fixes):

QA-1111 TA: Fix - “ID to Clipboard” option is missing in Advanced Search.
QA-1125 TA: Fix - After deleting any transcript, success message not showing.
BE-2022 TA: Fix - bad http request URLs in Edge deployments
QA-1140 TA: Fix - Project numbers going blank after filtering role in Users under Account.
QA-1081 TA: Fix - Proper translation should be there for mouse hovering on error.
BE-2021 TA: Fix - running out of nonces in some scenarios
BE-2027 TA: Fix - Sentry errors when sending to Glitchtip
QA-1184 TA: Fix - Shared transcript page getting refresh continuously(Public share)
QA-1141 TA: Fix - Sorting in Users under Account section is not working properly.
QA-1143 TA: Fix - The duration filter is not working as intended, having issues setting it up manually, and also not showing the transcripts.
QA-1167 TA: Fix - The owner is unable to log in to below mentioned IP addresses and the “Failed to fetch” error is being received.
QA-786 TA: Fix - Transcripts with no duration are not available in advanced search.
QA-1117 TA: Fix - Unable to edit “Users” when creating a new project.
QA-685 TA: Fix - Updated User name is not showing on the admin Speaker screen
QA-1116 TA: Fix - Walkthrough wizard not working properly, back button not working when user tries to go back from the transcripts.
BE-2013 TA: Fix - When inviting the user to the account, if we provide an invalid email it throws a 500 error and goes to the “Something went wrong” page
QA-1114 TA: Fix back button from transcript opened from Advanced Search results
BE-2084 Web Console: Fix - Audio is fresh-uploaded GREG experiment not playing
BE-2119 Web Console: Fix - certain Edge Configurations showing blank page instead
QA-1142 Web Console: Fix - Context dash filter by status is not working properly.
BE-1955 Web Console: Fix - In user Profile show permissions and not the Role
BE-2007 Web Console: Fix - Narrow window does not render transcript text in telephony call detail view
BE-2077 Web Console: Fix - Requests for GREG audio stuck in “pending” and then failing
BE-2087 Web Console: Fix - Showing Training mode even though account does not have Training enabled
BE-2053 Fix - Glitchtip cannot handle larger headers (cookies)
BE-1982 Fix - NPE found in new-billing-utility logs on voicegain
BE-2103 Fix - Queue info is not getting saved and/or returned from /sa/call API
BE-2083 Fix for negative duration of words
BE-655 Fix for onprem-cluster-deployment task stuck when pods are in unexpected status

Release 1.102.0

Key changes related to the core APIs

MRCP ASR now supports GARBAGE grammars (Nuance syntax)
Removal of Prompt Manager in Web Console - not used due to popularity of TTS
/sa/offline API is now available in beta - this is for Offline Transcription and Speech Analytics
Telephony Bot API supports A/B testing using multiple logic URLs
Improvements to accuracy of the whisper model

Key changes related to Transcribe APP

Cloud version uses whisper:small for English (it already used whisper:medium for foreign languages)
Generating just one Action Items table per entire transcript instead of one per Section.
Fixed errors in PDF and DOCX generation for some meetings

New or changed functionality in the Transcribe App:

BE-1595 TA: Improve the User Edit Dialog - project selection
BE-1699 TA: Better message if there are no other users on the Project
BE-1870 TA: Improve the User Delete dialog
BE-1871 TA: Make the Avatar icons larger
BE-1881 TA: Display Release Version in the app.
BE-1893 TA: Upload Zoom files using POST/data/s3 API
BE-1913 TA: Add option to copy session id to Clipboard
BE-1936 TA: Add browser client info to login request
BE-1975 TA Cloud: Enable whisper:small on English language
QA-1036 TA: Show allowed characters when entering a tag
QA-1041 TA: Added 365 days time limit for expiry of shared transcript.
QA-1089 TA: Added sorting by email to Users table

New or changed functionality in other platform components:

BE-1842 Web Console: Changes to API Security setting
BE-1843 Web Console: Better text in Settings -> Speech Recognition
BE-1844 Web Console: Hide Tools-> Prompt Manager menu option
BE-1882 Web Console: Display Release Version in the app.
BE-1896 Web Console: Switch Telephony Bot Call Audio+Transcript view from old /sa to new /sa/offline style
BE-1928 Web Console: Add configuration for multiple URLs to AIVR App Settings
BE-1929 Web Console: Add support for logic event in AIVR session
BE-506 Support removing objects with multiple versions in Google Storage
BE-857 Add key-match option to Query Terms in Advanced Search
BE-858 Add key field to Query Term in Advanced Search
BE-1150 Implement Offline SA Task in offline task project (Python)
BE-1406 Improvements to prvention of SQL injection in the Advanced Meeting Search API
BE-1514 Add to each of our API methods optional X-Request-ID header
BE-1550 Freeswitch: discontinue using modified mod_shout file
BE-1560 Update Freeswitch to use latest Debian 12
BE-1579 Collect session_duration for realtime ASR sessions
BE-1718 Implement POST /sa/offline/call/{callId}
BE-1719 Implement POST /sa/call
BE-1742 Mobile: Add extra parameter to /asr/transcribe/async request used in Microphone transcription
BE-1788 Switch AIVR recording processing from /sa to /sa/offline
BE-1789 AIVR to create /sa/call records if requested
BE-1816 As of Redis version 6.2.0, ZRANGEBYSCORE is regarded as deprecated
BE-1828 Add roles parameter to GET /user API method
BE-1848 (Demo-Voicebot) Convert voicebot demo details view to use /sa/offline
BE-1861 MRCP ASR: Add support for GARBAGE rule
BE-1862 Add initialPrompt to POST /asr/transcribe/async
BE-1867 For some internal APIs - ignore unknown fields, and not return 400 error, instead return X-Warning header
BE-1883 SSO: Display Release Version in authentication-client app.
BE-1884 Demo: Display Release Version in the app.
BE-1888 Add X-Request-ID to all API requests from all Web Apps
BE-1891 Support llm-svc rolling deployment – websocket sessions
BE-1903 Add version field to the AIVR session
BE-1904 Store the GCP service account key in edge cluster document in firestore
BE-1908 Add read-only offlineQueuePriority field to Account
BE-1910 Support POST /auth-svc/zendesk/jwt
BE-1912 Increase the size of concurrent websocket connections in asr-api
BE-1914 Add Client Params to the login API
BE-1919 Add numAudioChannels and numSpkChannels to POST /sa/call and GET
BE-1927 Add support for multiple call-back URL in AIVR App
BE-1940 Use default init prompt on Whisper if initPrompt is not set
BE-1942 In Meeting API, generate action items from entire meeting transcript
BE-1944 Add keySentencesByType to PUT /internal/asr/meeting/{meetingId}
BE-1956 Admin Tool: Add support for login using 2FA
BE-1959 Report license expiration time in edge-debugger
BE-1983 Improve the efficiency of relation extraction algorithm for IVR Prompt NLU to support long input
BE-2019 Modify sip settings of unimrcp server

Changes related to Integrity of Processing (fixes):

BE-1585 TA: Inspect all cases of using “dangerouslySetInnerHTML”
BE-1724 TA: Fix - SyntaxError: The string did not match the expected pattern, during fetchVersion in AppReloadModal
BE-1725 TA: Fix - TypeError: Failed to fetch version.json
BE-1727 TA: Fix - TypeError: c is not a function at handleDelete in components/ProjectsList/DeleteMultiSelectDialog
BE-1835 TA: Fix - Try Again button has no effect after a failed upload
BE-1841 TA: Fix - Unable to select audio files for upload on iPad
BE-1857 TA: Fix - Generated Avatar not working ok for single user account where use has no avatar picture uploaded
BE-1932 TA: Fix - Bad content security policy for GlitchTip
BE-1937 TA: Fix - After logging in, the home page initially loads but then suddenly goes blank
BE-1941 TA: Fix - TypeError: Cannot read properties of undefined (reading ‘includes’)
BE-1945 TA: Fix - TypeError: Cannot read properties of undefined (reading ‘value’) on Settings page
QA-1027 TA: Fix - Close Icon is missing on “Pair Voicegain Phone App” pop up
QA-1031 TA: Fix - User is able to set value above 365 days on Archival Text Redaction.
QA-1039 TA: Fix - User is able to set blank or invalid names as external speakers names.
QA-1044 TA: Fix - bad link in the guide
QA-1069 TA: Fix - Errors in PDF and DOCX generation for some meetings
QA-1083 TA: Fix - Inconsistent Upload success message on upload page
QA-1088 TA: Fix - Sorting indicators are broken for the Owned Projects and Shared Projects
QA-1092 TA: Fix - Projects with same names getting automatically selected in Advanced Search project filter.
QA-1093 TA: Fix - Player disappear when user clicks on anywhere near to the play button.
QA-1100 TA: Fix - Assigning a role for invited user should be a must required filed.
QA-1113 TA: Fix - Do not allow a tag with only underscores
QA-1114 TA: Fix back button from transcript opened from Advanced Search results
QA-641 TA: Fix - Account Owner should be irremovable for any the project under Account Users.
QA-800 TA: Improve behavior of resize on browser-share pop-up window
BE-1703 Web Console: Fix - Wrong error message in GREG
BE-1869 Web Console Edge: Fix - failed to decode request body: organization name "api_test" not found”
BE-1872 Web Console: Fix - TypeError: Cannot read properties of undefined (reading ‘ac’)
BE-1920 Web Console: Fix - Loss of filtering info in AIVR view
BE-1943 Web Console: Fix - context switcher does not correctly support duplicate-named project
BE-1962 Web Console: Fix - Error on Edge management page (properties of undefined)
QA-997 Web Console: Fix - Transcript is not getting skip by specified seconds as expected and onClick function is also not a function as in console tab.
QA-1086 Web Console: Only CMP users may modify CMP permissions
QA-1063 App Selector: Fix - Language selector should not be transparent.
QA-1066 SA: Fix - User is unable to change profile photo
BE-1478 Fix - Failed to execute ‘decodeAudioData’ on ‘BaseAudioContext’: Unable to decode audio data
BE-1840 Fix - PUT /sa/call/review/answers/{crAnswersId} Spec and Implementation are different
BE-1854 Fix - AIVR - prompt playback of audio from http url not working
BE-1902 Fix - llm-svc does not handle SIGTERM properly
BE-1907 Fix - uuid_vg_tap_ws at CLI not showing UUIDs to be started
BE-1925 Fix - AIVR playback of audio from an HTTPs URL stops after 30 seconds
BE-1999 Fix - Exception in case of many failed logins
BE-926 Fix - EqTerm is not properly handled in meeting search for some fields

Release 1.101.1

New or changed functionality:

BE-1877: Enable VAD on Whisper
BE-1886: Make node affinity configurable in offline-whisper
BE-1889: Deploy whisper model to the edge cluster without internet connection
BE-1938: Initial whisper prompt to ensure punctuation is generated

Release 1.101.0

IMPORTANT Note for Edge users: If you update from any prior release to 1.98.0 and you need to roll-back please contact Voicegain for support with the rollback process. This is because the compatibility setting on the Mongo DB has been changed in 1.97.0 and influxDB version has changed in 1.98.0

Key changes related to the core APIs

Revised User deletion logic - will retain deleted User info so that any remaining references can be resolved.
Added ability to enforce 2FA account-wide via a Web Console setting
Connected Web Console to Sentry service for error tracking
Improved session logs
Websocket version of the Telephony Bot API
Improved NLU model for understanding IVR prompts

Key changes related to Transcribe APP

Improved User deletion logic functionality.
Added LLM Settings to Account profile
Action Items tables now correctly rendered in PDF and DOCX
Upload audio data directly to storage without sending data through data API service
Improved tables with lists of Shared transcripts (User and Admin view)

New or changed functionality in the Transcribe App:

BE-1189 TA: Add option for Admins to see what others have shared
BE-1271 TA: Support a workflow for the Admin to delete a user account and take over the user’s project
BE-1536 TA: Improved Splash page after login which shows the different stages
BE-1646 TA: If there are no devices we show 2 options: download and phone app setup
BE-1708 TA: On All Shares table, added sorting and filter on the Creator column, and filters on the Scope and Expires columns
BE-1709 TA: In account users show Own Projects and Shared Projects columns in place of the current single Projects column
BE-1713 TA: Modify the PDF output to render markdown tables
BE-1714 TA: Modify the DOCX output to render markdown tables
BE-1721 TA: Improved error message on the Zoom directory upload
BE-1733 TA: Generate avatar based on user name if they have not uploaded a picture for avatar
BE-1741 TA: Add extra parameter to /asr/transcribe/async request used in Microphone transcription
BE-1744 TA: Upload files using the POST /data/s3 API
BE-1807 TA: Add LLM Settings to Account profile
BE-1810 TA: More functional user delete
BE-1818 TA: Show info of users that have been deleted
BE-1823 TA: Add Polish language transcription
BE-1834 TA: Remove “Sync From Cloud” button
QA-1009 TA: Increase max number of transcripts shown on home page from 100 to 250
QA-1010 TA: Show the Upgrade button only to the Owner role
QA-1017 TA: Better error message in case of an error resetting password
QA-1021 TA: When user is deleted all user sessions will be invalidated.

New or changed functionality in other platform components:

BE-1194 Advanced Search: Add new column bucket to the meeting_session table
BE-1579 Collect session_duration for realtime ASR sessions
BE-1663 Web Console: When deleting Users in the Cloud and if there are Edge deployments warn that also users on the Edge will be deleted
BE-1687 Websocket version of AIVR API: support question.audioResponse.streaming
BE-1693 Support Zoom Breakout Rooms in Meeting Join
BE-1695 New /sa/call/search API
BE-1697 New /sa/call/search/fields API
BE-1691 Web Console: Add internal refresh to the page with Edge Deployment details
BE-1705 Web Console: Add ability to enforce 2FA account-wide
BE-1706 Add to Account API a field that controls 2FA enforcement
BE-1707 Web Console: Better error message in case of creating a GREG experiment with duplicate name
BE-1715 Support model_name measurement for offline sessions (offline transcribe, offline meeting, offline SA)
BE-1716 Add session ID to all log messages in offline-task
BE-1722 Web Console: Improved inactivity timeout processing
BE-1728 Update code to use the latest Azure OpenAI Service preview API
BE-1732 Web Console: Connect to Sentry
BE-1739 Connect authentication-client to Sentry
BE-1743 Add llmSettings to /cluster/ API
BE-1809 Add deleteUserContexts parameter to DELETE /user
BE-1817 New inclDeleted parameter on GET /user API method
BE-1822 Support inclDeleted for GET /account/uuid/users
BE-814 Improved NLU model for IVR prompts
BE-978 Removed contextId and fromAllContexts parameters from the Advanced Meeting Search API method
QA-1046 Demo: Improved error message in case of internal issues.

Changes related to Integrity of Processing (fixes):

BE-1629 TA: Fix - It is impossible to share a project with an Owner of the account
BE-1734 TA: Fix - Sorting of devices by date broken if any device is deleted
BE-1802 TA: Fix - Search for creator by name in Advanced Search filter
QA-1012 TA: Fix - Project setting- Save button enabled when there is no change
QA-1022 TA: Fix - In some rare cases Voice Signature page is showing white page.
QA-887 TA: Fix - Walk through wizard should get automatically initiated when new user logs in for the first time.
QA-929 TA: Fix - Project (number) information is confusing
QA-967 TA: Fix - User is able to add invalid keywords under project setting.
BE-1672 Web Console: Fix - Sliders too small at console.ascalon.ai/specific/meetings
QA-992 Web Console: Fix - User is able to play the audio when no voice is selected while creating a phone app.
QA-993 Web Console: Fix - After clicking on view button of call session, it’s showing blank page.
QA-997 Web Console: Fix - Transcript is not getting skip by specified seconds as expected and onClick function is also not a function as in console tab.
QA-999 Web Console: Fix - When User click on Cancel button of add context it should be close.
BE-1735 Mobile App: Fix - order of projects in the list
BE-1736 Mobile App: Fix - My project returns transcripts from other projects
BE-1812 AIVR: Fix - getting MATCH but no recognized utterance
BE-1831 ASR API: Fix - ValueError: Invalid value for type (), must be one of [‘ner’, ‘regex’]
QA-1005 SA App: Fix - use the correct time format enum values

Release 1.100.0

New functionality in the Transcribe App:

BE-1559 TA: Automatically update the currently installed Zoom App version
BE-1582 TA: Add “Mine Shared All” selector to the home page
BE-1625 TA: Add a logout confirmation dialog.
BE-1642 TA: Show avatars on Account Users table
BE-1659 TA: Support action items using GPT for cloud Transcribe App
BE-1665 TA: Support action items using custom LLM for Edge
BE-1669 TA: Advance Search - Type filters populated dynamically from audio_src in fields api.
BE-1671 TA: Hide the project settings for the Key Items

New functionality in other platform components:

BE-1315 Move rexConfig from xml to properties file in Rex
BE-1348 After password reset in AuthenticationClient provide link(s) to the correct App(s) for a given user
BE-1527 Switch to yaml rex configuration file on K8s
BE-1567 Web Console Edge: Add splash page after login and before home page shows up
BE-1578 Changes to get app.voicegain.ai score A on https://securityheaders.com
BE-1593 Web Console: Remove Clouds from ACP Header on Cloud
BE-1628 Meeting Join API - add parameter to control which js script is run (zoom, teams, etc)
BE-1638 Changes to the Voicebot Demo Instructions page
BE-1648 Web Console: Disable old MRCP Experiment Analyze - the new will become the only one

Changes related to Integrity of Processing (fixes):

BE-1596 TA: Fix - Zoom Device Approvals pop up on wrong accounts
BE-1629 TA: Fix - It is impossible to share a project with an Owner of the account
BE-1649 TA: Fix - When the delete button is clicked in advanced search, it triggers an incorrect API, leading to inaccurate transcript result, Furthermore, post-deletion, filtered and searched data fail to display accurately.
BE-1652 TA: Fix - Reload button does not work on Firefox
BE-1664 TA: Fix - deleting a project should also remove its meetings from databases
BE-1702 TA: Fix - Advanced Search page goes blank with TypeError: Cannot read properties of undefined (reading ‘split’)
QA-822 TA Edge: Fix - User (Role) is able to access Account settings.
QA-853 TA: Fix - No options to select under “creator” filter in advanced search.
QA-856 TA: Fix - Alert msg of “maximum allowed minutes exceeded” is not translating in other languages.
QA-888 TA: Fix - In Advanced search deleted transcript are not getting removed from list, unless we refresh the page.
QA-892 TA: Fix - partial Translation is there while uploading a file
QA-903 TA: Fix - There should be proper translation on the share pop-up page.
QA-913 TA: Fix - Walk through wizard gets hidden after few steps, when left menu is locked.
QA-920 TA: Fix - For a single project, the move should be disabled
QA-931 TA: Fix - User is able to de-select all filters on home page for shared.
QA-938 TA: Fix - Different loading screens when deleting a single share in compared to deleting multiple shares.
QA-940 TA: Fix - User is unable to “generate “ the JWT token.
QA-944 TA: Fix - Owner should not be allowed to delete all of its project if any user have shared any project with him.
QA-946 TA: Fix - Unable to Invite user after we edit a user in account section.
QA-947 TA: Fix - Voicegain splash screen showing when project user adds other users to a project.
QA-949 TA: Fix - Unable to close the dialog pop-up for added phrase.
QA-950 TA: Fix - Project setting- save button overlaps with text when Spanish language is selected.
QA-953 TA: Fix - Account with only user access can invite other account.
QA-961 TA: Fix - While creating the first project(new project wizards) Word cloud toggle not showing properly.
QA-962 TA: Fix - User with Admin role is not able to update address in profile section
QA-963 TA: Fix - Language reminder- Close icon not showing properly cutting with boundary of the pop-up.
QA-964 TA: Fix - During Signup user given company name and title, after login title missing under profile section
QA-970 TA: Fix - Date filter is not working on advanced search page.
QA-972 TA: Fix - Invited User is able to change “May Download” setting in account.
QA-973 TA: Fix - Language filter under advanced search should show full language name
QA-985 TA: Fix - Save button not working properly under profile setting.
BE-1512 Fix - Occasionally k8s lost gpu and we need to restart nvidia-device-plugin-daemonset
BE-1591 Web Console: Fix - Audio display in Meeting Detail page messed up
BE-1641 Web Console: Fix - GREG is not showing recognition results
BE-1644 Fix - Something wrong with keyword advanced search
BE-1650 Fix - accounts are getting revoked on QA even though the setting is 0 which means do not revoke
BE-1654 Fix - Misleading error response of POST API asr/meeting/join
BE-1668 Fix - Zoom Meeting join api cannot handle a case where the waiting room is disabled
BE-1674 Fix - NullPointerException from JoinedMeetingEvent
BE-1675 Fix - NullPointerException is thrown from ascalon-cleanup
QA-748 Demo: Fix - Double confirmation is required to browse audio file in the upload demo file
QA-797 Web Console: Fix - On Websocket Details on live broadcasting getting the blank page
QA-894 Web Console: Fix - Page is not responsive enough for the 90% zoom, all the pop-up are vibrating.
QA-943 Web Console: Fix - Unable to set “Disable inactive users” as Not Enabled.
QA-983 Web Console: Fix - After reset password- Web console button is not clickable

Release 1.99.0

New functionality in the Transcribe App:

BE-1258 TA: Add a filter on creator in Advanced Search
BE-1472 TA: Better order of columns in the transcript table on home page
BE-1532 TA: Record in clientSideProperties when a user downloads Zoom Meeting Assistant
BE-1582 TA: Add “Mine Shared All” selector to the home page

New functionality in other platform components:

BE-1300 Meeting Search API: new ctxSelect parameter with values “mine”, “mineAndShared”, “all”
BE-1301 Meeting Search Fields API: new ctxSelect parameter with values “mine”, “mineAndShared”, “all”
BE-1361 Support tempCode in the AIVR API
BE-1408 Modify the Password Reset API to return login URLs for all the Apps for which the password works
BE-1446 Return 400 Bad Request rather then 500 Internal Server Error if the query parameters in Meeting Search API are missing or invalid
BE-1458 Web Console Edge: Add check for email relay and do not offer emailed OTP if that email relay is not configured
BE-1488 Configure Edge to have GlitchTip
BE-1499 Add API that Revokes a user and transfers all projects to the Admin or some other designated User
BE-1517 Convert /sa/call API to use Postgres instead of Firestore to store data
BE-1522 API to email link to the results of Voicebot Demo session
BE-1524 Deploy GlitchTip using onprem-cluster-deployment
BE-1526 Reduce silence padding of Azure TTS prompts to minimum
BE-1533 Add persist setting to AIVR app API
BE-1534 Web Console: Add persist setting for Telephony Bot App
BE-1538 Hookup Demo App to Sentry
BE-1541 Cache account data daily using the /recache api
BE-1544 Web Console: Do not allow smaller than 30 days values of “Disable inactive users after days”
BE-1545 Web Console: Add a close (x) to the notice about 2FA setup
BE-1546 Web Console: Add a spinner when saving account settings
BE-1565 For accounts that are INVOICE and do not have billingAccountId we should not submit anything to fusebill
BE-1572 AIVR Callback API method to interrupt prompt being played
BE-1575 Utility that copies all calls - reads using Firestore Calls DAO and writes using Postgres Calls DAO
BE-1580 Add jingleUrl field to the AIVR App
BE-1581 Web Console: Add ability to configure Jingle and percolator URLs in AIVR APP
BE-1588 Add percolatorUrl field to the AIVR App API
BE-1589 Use the AIVR App Jingle in Freeswitch script
BE-1592 Web Console: Make it possible to disable the “Disable inactive users after days” feature
BE-1594 Web Console: Make the Account Settings card pretty
BE-1599 Demo: Support copy paste of the Code
BE-1604 Demo: Change tooltips
BE-1615 When creating Users using POST /user, if we exceed max users on account we should not return 500 code but 429

Changes related to Integrity of Processing (fixes):

BE-1621 TA: Fix - User Profile Setting Save button stays permanently disabled after saving once
BE-1629 TA: Fix - It is impossible to share a project with an Owner of the account
QA-641 TA: Fix - Project Owner should be irremovable for any the project under Account Users.
QA-760 TA: Fix - The re-upload menu again shows for the normal uploaded file, it should only show for the zoom uploaded file.
QA-822 TA Edge: Fix - User (Role) is able to access and edit Account settings.
QA-843 TA: Fix - When the user tries to share the transcript as a public share and delete the transcript from the owner account, something went wrong error is showing when login in from the sign-in link on the share.
QA-860 TA: Fix - Mouse hovering on error, message information should show
QA-863 TA: Fix - “May Download” checkbox getting automatically disable after adding/removing a project for user (role).
QA-875 TA: Fix - After processing URL upload blank space showing on home page.
QA-881 TA: Fix - Clicking on the filter button on advanced search page, showing blank page.
QA-885 TA: Fix - Voicegain splash screen showing even when deleting a transcript.
QA-890 TA: Fix - Due to new languages added, profanity maskimg menu not showing properly on 100% zoom screen.
QA-891 TA: Fix - In Profile deleted Shares are not getting removed, unless we refresh the page.
QA-899 TA: Fix - Expired shared transcripts showing expiry date in negative.
QA-906 TA: Fix - Various problems with saving edited user data from Account Settings
QA-917 TA: Fix - Text-redaction and Archival Text-redaction crashes the app
QA-922 TA: Fix - Blank page showing when user clicks in Profile on the “Archival text reduction”.
QA-923 TA: Fix - If no Download permission selected on user invite, getting wrong permission error
QA-925 TA: Fix - Getting blank screen on Users page under account section.
QA-932 TA: Fix - Voicegain splash screen showing when user save the mic capture recording.
BE-1484 Web Console: Fix - On Live Transcribe: Websocket Delay resets to 0 upon page refresh
BE-1485 Web Console: Fix - Misaligned UI on the Live Transcribe Websockets
BE-1576 Web Console: Fix Content Security Policy to the cc app iFrame
BE-1577 Web Console: Fix - /restricted/page API returns 500 Internal Server Error
BE-1586 Web Console: Fix - Wrong request to retrieve AIVR sessions
BE-1603 Fix - Microphone demo needs to provide also completeTimeout
BE-1624 Fix - User REVOKE can be undone by a user by doing a simple Password Reset
QA-847 Web Console: Fix - After uploading a file, the label field should be automatically filled.
QA-889 Web Console: Fix - Walk through wizard should get automatically initiated when new user logs in for the first time.

Release 1.98.0

New functionality in the Transcribe App:

BE-1395 TA: Improvements on the Share dialog
BE-1494 TA: Add Splash page on login
BE-1507 TA: Support French, Dutch, Portuguese, Italian languages using the whisper:medium model (Cloud only)
BE-1508 TA: Hide the Key-Items configuration in settings of Projects with language other than English

New functionality in other platform components:

BE-1006 Collect lastConnectedEpoch time for device
BE-1226 Send logs and errors from Voicegain Flutter app to Sentry
BE-1358 Web Console: Inactive Users are disabled after 90 days of inactivity - add ability to set disableInactiveUsersDays value
BE-1365 Remove default sort by meeting_id from the GET and POST /asr/meeting/search
BE-1371 Extend default persistence of Telephony Bot calls to 42 days
BE-1380 In the SSO UI add ability to mark the browser safe (i.e. no longer requiring TOTP)
BE-1404 Add new field to account API - disableInactiveUsersDays
BE-1431 Support Influxdb V2 on Edge
BE-1438 Edge: Write sizes of minio and mongo backups as influxDb measurements
BE-1457 Web Console: Show time in the deployment history in the time zone of the browser and in a nicer format
BE-1460 Add a ‘generic’ joined meeting event and annotate the zoom meeting with such events for various steps performed
BE-1461 Add joinedMeetingEvents to the Meeting API poll method response
BE-1462 Modify login API to return information if emailing OTP is possible
BE-1470 Add image to the url/QR that is used in TOTP setup
BE-1474 Modify the existing login API to take secureBrowserToken
BE-1475 Implement internal API to write sizes of backups to influxDB
BE-1481 Web Console: Turn the nagging 2FA dialog to a message that auto disappears
BE-1482 Web Console: Improvements to the 2FA setup dialog
BE-1483 Modify the share APIs to track usage of shares
BE-1493 Show number of currently open MRCP and RTSP sessions in the unimrcp log
BE-1506 Web Console: Allow for selection of multiple languages plus base, small, and medium whisper model
BE-1513 Deploy prometheus to edge cluster using onprem-cluster-deployment task
BE-1516 Autoscaling Offline transcription on Edge deployed on GCP
BE-1522 API to email link to the results of Voicebot Demo session
BE-1523 Disable returning meeting topic generation for languages other than English and Spanish
BE-1525 Change max value of poll.persist to an equivalent of 365 days
BE-1551 Increase Project limit to 1000 and change the corresponding error from 500 to 429
BE-502 Modify User PUT API and add ability to add secureBrowserToken
BE-776 Upgrade unimrcp to 1.8.0
QA-563 Admin Tool: Remove Register page

Changes related to Integrity of Processing (fixes):

BE-1440 TA: Fix - Different Users not able to create Projects with the same name.
BE-1549 TA: Fix - For Zoom dir upload the size check should be performed only on files that actually are going to be uploaded
QA-723 TA: Fix - First time Login- All the links should be hyperlinked.
QA-760 TA: Fix - The re-upload menu again shows for the normal uploaded file, it should only show for the zoom uploaded file.
QA-828 TA: Fix - New User First Login - Zoom App page is coming blank
QA-834 TA: Fix - Hint message is missing below Tag input field
QA-864 TA: Fix - Error after new login if the Project user used last has been deleted
QA-865 TA: Fixes in the speaker number range selection UI
QA-887 TA: Fix - Walk through wizard should get automatically initiated when new user logs in for the first time.
BE-1278 Fix - If no sort order in specified then the advanced text search results should be ordered by rank
BE-1334 Fix - results from the GET /asr/meeting/search API are not sorted by rank if no sorting is specified
BE-1436 Fix - Recognition doesn’t work in AIVR sessions if influxDB is down
BE-1490 Fix - AIVR should start recognition only after the non-bargineable text prompt
BE-1563 Fix - Missing <instance> value in the result of builtin:speech/transcribe (and all lvoc “grammars”)
QA-847 Web Console: Fix - After uploading a file, the label field should be automatically filled.

Release 1.97.3

Updates:

BE-1455 Upgrade cloud function runtime from python 3.7 to 3.9

Release 1.97.2

Changes related to Integrity of Processing (fixes):

BE-1510 Fix - SnippetAnnotations is not working

Release 1.97.1

Changes related to Integrity of Processing (fixes):

BE-1412 Fix - uhhuh recognition with low confidence messing up the flow (early barge-in)
BE-1443 Fix endpointing of low confidence recognitions in Telephony Bot API
BE-1473 Fix - NPE in Meeting Search if any account context does not have type set
BE-1476 TA: Fix - Something Went Wrong page not loading correctly on Edge
BE-1477 Fix - Invalid value: null error when retrieving transcript in Web Console
BE-1479 TA: Fix - missingKey translation “Uploaded File” logged in a loop
BE-1487 Web Console: Fix - Unable to sign-up without providing Company Name
QA-858 Web Console: Fix - Showing no transcript when click on the view for any transcription.
QA-859 Demo: Fix - Getting white screen in demo when trying to upload a file or trying to do doing mic capture

Release 1.97.0

IMPORTANT Note for Edge users: If you update from any prior release to 1.97.0 and you need to roll-back please contact Voicegain for support with the rollback process. This is because the compatibility setting on the Mongo DB has been changed in 1.97.0.

New functionality in the Transcribe App:

BE-1186 TA: Meeting search fields API returns intermediate values
BE-1377 TA: In Advanced Search filters use the intermediate values returned from the fields API
QA-421 TA: Improved inactivity timeout functionality
QA-660 TA: Transcript caption popup window now is resizable in full-screen video player

New functionality in other platform components:

BE-1249 Added logicConnectMethod field to AIVR App API
BE-1353 Update Node version app-selector: 16.10.0 → 18
BE-1355 Web Console: Show the SHA and description of the currently deployed Edge configuration
BE-1356 Web Console: Show the history of deployed Edge configurations (incl SHA sums)
BE-1359 Web Console: Make password expire time configurable
BE-1360 Web Console: Make number of old remembered passwords configurable
BE-1367 Migration from InfluxDB v1.8.10 to v2.7.x
BE-1376 Add joinedMeetingEvent(s) field to the /asr/meeting API for meetings that were joined using meeting bot
BE-1379 Web Console: Add ability to set logicConnectMethod for AIVR App
BE-1397 In account API make number of old remembered passwords configurable
BE-1398 In account API - new min, max, and default values of inactivityLogoutThreshold
BE-1399 Add passwordExpirationDays to the Account API
BE-1400 In MRCP, support defining large vocabulary grammar using buildin grammar in DEFINE-GRAMMAR request
BE-1403 Web Console: Added option to control which Edge configurations will be shown
BE-1405 Add deploymentHistory field to /cluster API
BE-1422 A utility that creates user/org/bucket in InfluxDBv2 for each existing account in the db
BE-1434 Support storing orgId and userId under account.influxDB
BE-1435 Support hostAliases in offline task
BE-1441 Support acoustic model with only model config file and a reference to another model
BE-1442 Create a telephony bot specific acoustic model configuration in REX with new sensitivity setting
BE-1444 Log any occurrence of “Something Went Wrong” page to Sentry
BE-473 2FA for Web Console
BE-503 Web Console: In the user profile add ability to configure 2FA
BE-504 In Authentication Client modify login to handle cases if MFA is enabled for a user

Changes related to Integrity of Processing (fixes):

BE-1252 TA: Fix - After Submit on Zoom Upload still staying on the same page
BE-1383 TA: Fix - The User avatar is currently saved to the currently selected context - we need to save it to another bucket
BE-1385 TA: Fix - Password reset link incorrect on some Edge deployments
QA-509 TA: Fix - Incorrect behavior if User tries to use Share “within account” while not being logged in
QA-594 TA: Fix - On uploading the zoom folder getting the status as error
QA-703 TA: Fix - Shares is not showing under My Shares (on accounts with many transcripts)
QA-709 TA: Fix - Filter by Phase/Status not working in advanced search.
QA-756 TA: Fix - Zoom App Page is not loading for some new users
QA-799 TA: Fix - For new users “Something went wrong” error page shown instead of a First Project Wizard
QA-802 TA: Fix - Save button is not getting enabled after updating the settings.
QA-809 TA: Fix - Advanced Search Filter is not working on Status Filter
QA-811 TA: Fix - Advanced search: Searching for the error transcripts filter not working.
QA-820 TA: In Advanced Search now able to enter duration manually (in addition to slider)
QA-829 TA: Fix - For new User Search Page is coming blank
QA-833 TA: Fix - After changing the profile picture “Hide Project Language Reminder” checkbox automatically gets unchecked.
BE-1363 Fix - Hints with misspellings sometimes do not work correctly in real-time mode
BE-1387 Fix - VAD classifies Beep as Speech
BE-1389 Fix - Limit of 3 pending pairing devices should be per User and not per Account
BE-1391 Fix - Edge data backup cronjob has wrong minio password
BE-1401 Fix - Some vars are ignored in AIVR callback
BE-1436 Fix - Recognition doesn’t work in AIVR sessions if influxDB is down
BE-1447 Fix - MaxAskExceedException in llm-svc is not captured
QA-796 Web Console: Fix - Signup error message in XML format
QA-823 Web Console: Fix - Ascending DNIS and ANI sorting not working for call sessions table.
QA-824 Web Console: Fix - Delete functionality is not working Properly for 90% page zoom.

Release 1.96.0

New functionality in the Transcribe App:

BE-1158 TA: In Advanced Search show Headlines in the list of search results
BE-1172 TA: Identify currently paired Zoom Meeting Assistant
BE-1193 TA: In Advanced Search - Add filter on transcript language
BE-1258 TA: Add a filter on creator in Advanced Search
QA-350 TA: Upload button is now disabled when max allowed minutes are exceeded for the basic plan.
QA-671 TA: Revised search location of the Home page
QA-726 TA: Voice signature status is shown in specific colors like green for “Ready” and blue for processing.
QA-729 TA: Added special page for 500 errors
QA-773 TA: Tags have their own specific alert message on Editing (Adding and Deleting).
QA-778 TA: Improved login behavior if accessing a meeting shared within Account

New functionality in other platform components:

BE-1084 AVIR: Added Telephony Bot Websocket API
BE-1175 Upload large files via a pre-signed URL to S3 (including Google Storage in S3 mode)
BE-1192 Support the new lang column in the GET fields method for search
BE-1222 Add intermediateValues field to the GET /asr/meeting/search/fields API
BE-1285 Add headline to response from the Advanced Meeting Search API
BE-1325 Rewritten built-in grammars to avoid repeat when it’s not in an scope
BE-1350 Update Node version for customer-portal: 14.21.3 -> 18
BE-1351 Web Console: Comment out Language Model which is no longer needed for Audio Sending Daemon
BE-1361 Support tempCode in the AIVR API
BE-1362 Configure ingress for wss://host:port/n/ws/aivr/uuid/session/uuid
BE-1381 In MRCP ASR - in Large Vocab Transcription return combination of 2 language recognition
BE-1384 Web Console: Add filtering AIVR call sessions by App Name
QA-779 Web Console: Better error messages if phone number not found
QA-813 Web Console: Added TTML download option

Changes related to Integrity of Processing (fixes):

BE-1333 TA: Fix - Mute/Unmute gets confused if we press the mute/unmute button too fast in a seqeunce several times
BE-1336 TA: Fix - Refresh icon gets squished when we open search box
QA-669 TA: Fix - Various issues on Advanced search feature
QA-672 TA: Fix - Unable to choose the folder directory in Zoom upload
QA-710 TA: Fix - Page behavior is unusual after URL upload.
QA-718 TA: Fix -Search for transcript on the home page is not working if searching by transcript name
QA-752 TA: Fix - Owner account should not show while creating a project and adding users to the project.
QA-768 TA: Fix - Advanced search page is going blank.
QA-770 TA: Fix - Filters are getting reset, when user open any transcript through advanced search.
QA-772 TA: Fix - Load more button the Advance search not working properly it respond after 3 to 4 clicks
QA-774 TA: Fix - Save button shouldn’t get enable if inactivity timeout field is empty.
QA-780 TA: Fix - Tags, Speakers, Participants filters associated with zoom upload are not searchable in Advanced search.
QA-781 TA: Fix - Sometimes Word Cloud gets broken for some random transcripts.
QA-782 TA: Fix - All 5 languages should be mentioned under language filter.
QA-792 TA: Fix - “May Download” checkbox getting automatically enable after adding the project in user setting
QA-794 TA: Fix - No error or alert msg on login page, when login with incorrect password.
BE-1340 Multiple 3rd-party Vulnerabilities fixed
BE-1342 Web Console: Fix - Error messages mixed up after purchasing a phone number
BE-1349 Fix: asr-api doesn’t have to check if a websocket exists for known topics
BE-1370 Web Console: Fix - Right channel of the Telephony Bot recording is not getting shown
QA-767 Web Console: Fix - Shortcut keys are not working properly.

Release 1.95.1

Changes:

BE-1343 Add property builtin.grammar.output.flavor to dynamic grammar

Release 1.95.0

New functionality in the Transcribe App:

BE-1255 TA: After new content is loaded by clicking “Load More” the page should scroll to the same place as before
BE-1256 TA: Do not truncate meeting names if not needed
BE-1274 TA: Check sizes of all Zoom folder files before starting upload and not during
BE-1276 TA: Add our own controls over the small video window of Shaka Player
QA-719 TA: Now, after search the page scrolls past “Get Started” and “Actions” to results
QA-720 TA: Share link now does proper redirect back to the share link if login is required

New functionality in other platform components:

BE-996 App Selector: Tweaks on the Signup page
BE-1143 Upgrade from GCP container registry to artifact registry
BE-1176 A method to stream words to ml-svc and get SA results
BE-1197 Web Console: Show an error if phone number purchase fails
BE-1209 Web Console: Show any error that occurs when creating a context
BE-1233 Web Console: In edge deployments view make the number of rows per page configurable
BE-1262 add /opt/voicegain/bin and /home/voicegain mounts to a k8s containe
BE-1288 Support semicolons as query string separators in dynamic-grammar
BE-1289 Support returning SWI_literal in dynamic-grammar
BE-1290 Updated Grafana in edge deployment task
BE-1291 Configure Sentry on Edge
BE-1292 Create security policy for sentry.io on Transcribe App
BE-1318 Web Console: Add more voices to AIVR App selection
BE-1321 Deploy llm-svc to qa and prod
BE-1323 Report llm-svc gpt cost for each session
BE-1326 Support both builtin grammar:digits and grammar:digit

Changes related to Integrity of Processing (fixes):

BE-1221 TA: Fix - Uncaught TypeError: Cannot read properties of null
BE-1257 TA: Fix - For Duration the sorting toggle is placed weird
BE-1259 TA: Fix - Keywords and tags displayed are missing spaces between them
BE-1261 TA: Fix -Show correct color for each project in Project filter
BE-1277 TA: Fix - Sometimes the Zoom Meeting Assistant page show stale value of the installed version
BE-1299 TA: Fix - When searching for text in meetings on home page, the API request should not contain any sort parameters
BE-1304 TA: Fix - POST /auth-svc/auth/login/openid fails if the User account exists but is only in the CREATED state
BE-1313 TA: Fix - weird behavior of Project search
BE-1314 TA: Fix - Weird logic for showing Move button on multiple selects
BE-1317 TA: Fix - On Edge the option to add users to project is not visible to the Project Owner
QA-651 TA: Fix - Date formats should not get translated in other languages.
QA-657 TA: Fix - File submit fails if we change project after upload and before submit.
QA-677 TA: Fix - Show current month in the right pane of the calendar
QA-698 TA: Fix - Project search by name is not working properly.
QA-699 TA: Fix - At the time save Transcript if we change the project then getting error
QA-700 TA: Fix - User is able to update the email in update payment but updated mail is not showing after saved
QA-701 TA: Fix - Add user- There should be words limit for entering the name
QA-705 TA: Fix - Tag for the voice signature showing like a single word not showing any gap between them.
QA-706 TA: Fix - Overlapping text in Download option when Spanish or German language is selected
QA-716 TA: Fix - If the chat is long then White blank screen showing after the Video
QA-717 TA: Fix - On large Chat video view -closed captioning, Video minimize buttons is not working
QA-721 TA: Fix - My Shares page stuck on loading after we Edit a share.
QA-725 TA: Fix - User should only able to delete the voice signature by clicking on the delete icon.
QA-727 TA: Fix - mouse hover on the Regenerate button showing in English when Hindi language is selected.
QA-732 TA: Fix - There should be a limit for max allowed char for the project name.
QA-733 TA: Fix - My shares table - tags should be separated.
QA-735 TA: Fix - Restart when adding a new voice signature is not working properly.
QA-736 TA: Fix - On the archival text reduction page, unable to save the updated time as the save button not enabled when changing days.
QA-741 TA: Fix - On Advance search Project check box is not showing similar for all project
QA-743 TA: Fix - On Changing the setting of “Start of the calendar week” from Sunday to Monday and vice versa getting the error of enter valid data
QA-744 TA: Fix - Invited user is able to move the transcript of Admin project But the Admin is not able to move the transcribe to User project and also the error message is not showing on the Frontend
QA-749 TA: Fix - Project creation page-users page is missing
QA-753 TA: Fix - Re-Upload option is showing for Recording and Browser.
QA-757 TA: Fix - On navigating back from the browser arrow from the large video mode then left menu disappear from the screen
BE-1200 Admin Tool: Fix - Duplicated Search button label
BE-1280 Fix - RexServerLauncher bean should start before all controllers
BE-1296 Fix - Rate-limit cleanup cronjob is not created on CHD environment
BE-1305 Fix - meta..text should also include text in its reference rules
BE-1316 Fix - audio-server fails to start on Edge
BE-1319 Fix - audio-server fails to compute duration of down-sampled audio generated by maryTTS
BE-1327 Fix - dtmf currency grammar doesn’t work
BE-1329 Fix - Missing Content-Length header in the response of GET /private/synthesis
BE-716 Fix - Grafana image rendering in version 8
QA-708 Web Console: Fix - Incorrect pop-up msg when updating company address in account settings.

Release 1.94.1

Changes:

BE-1282 Fix - offline task can’t process 7.1 channel audio

Release 1.94.0

New functionality in the Transcribe App:

BE-1054 TA: Avatar image can now be edited (cropped)
BE-1104 TA: Better reporting of file size errors on upload (413 error)
BE-1167 TA: Show Project settings only to Project owners and Admins
BE-1168 TA: Add user Profile Setting: “Show advanced Project settings”
BE-1170 TA: Any text redaction Settings (Account, Project) visible only to Account Owner and Admins
BE-1171 TA: Menu actions in the Advanced Search results list work now the same as on the Home Page
BE-1174 TA: Added “Re-upload” that opens Zoom Upload form
BE-1190 TA: Show countdown on the modal that shows the QR code.
BE-1196 TA: Prompt user to reload if there is a new version (also invalidate any open session)
BE-1234 TA: Do not show Analytics settings on a Project unless user has enabled advanced settings
BE-1236 TA: Upgrade to Node version 18
BE-1254 TA: For Admins the “Show advanced Project settings” is now enabled by default
BE-1264 TA: Added more logs around the login process
QA-602 TA: Cleaned up the Inactivity Timeout setting
QA-619 TA: Remove settings in Shaka Player
QA-642 TA: Similar Date formats are now together in Settings under Account in Profile menu.

New functionality in other platform components:

BE-984 Update helm to the latest version
BE-1114 Move feature extraction to Triton to reduce CPU load in ASR
BE-1115 New Date Format enum for Account and Context
BE-1127 Move from tfs to triton for VAD models
BE-1144 Extract audio features using Triton
BE-1191 Meeting search API: Add new field: LANG - to store the main language of the transcript
BE-1203 End-to-end flow for meeting recorder using Puppeteer (Webex)
BE-1213 Hardened core ASR when the model is not available
BE-1219 Use UTF-8 encoding in telephony bot callback requests
BE-1223 Web Console: Add column sorting on the phone management page plus variable page size
BE-1224 Add debug meetingPlatform value to POST /asr/meeting/join
BE-1240 Support prompt response in llm-svc (in addition to question, transfer, disconnect)
BE-1246 Support custom mrcp configuration using helm chart
QA-654 Demo: Show message confirming copy to clipboard

Changes related to Integrity of Processing (fixes):

BE-875 TA: Fix -Warning about a too large file is not being shown
BE-1105 TA: Fix - Submit button gets enabled too early while uploading zoom recording folders
BE-1128 TA: Fix - “walkThroughWizardSeen” gets set to True even if user has not done the Wizard
BE-1156 TA: Fix - In Advanced Search Filters should be remembered
BE-1157 TA: Fix - In Advanced Search show Project icon as the first column
BE-1169 TA: Fix - Show error message when hovering over the Error in Transcript list (was broken)
BE-1180 TA: Fix - When SSO user gets logged out from TA they see normal login page for a few seconds
BE-1182 TA: Fix - Weird behavior after clicking back button on detail page opened from Advanced search
BE-1217 TA: Fix - Advanced Search API should return al the values if body is not passed.
BE-1220 TA: Fix - Missing translations
BE-1265 TA: Fix - First Project wizard not running is new user was invited to some projects
BE-1269 TA: Fix - On a new account the My Shares page hangs forever with a spinner
BE-1270 TA: Fix - GET /asr/meeting sharedBy returns meetings for a brand new user that has not shared anything
BE-1272 TA: Fix - Getting incorrect error “cannot transcribe audio from video file” while uploading zoom folder
BE-1273 TA: Fix - Checking for file size in Zoom Upload is broken (looks at files that will not be uploaded)
QA-537 TA: Fix - Current Project is not picking correctly while move
QA-613 TA: Fix - On uploading zoom folder- chat not uploaded correctly
QA-643 TA: Fix - Search is not working on Homepage
QA-644 TA: Fix - Do not show tag edit option in Shared transcripts
QA-655 TA: Fix - Submit button shouldn’t get enable until all selected files are uploaded.
QA-670 TA: Fix - Right outline of the box is missing the speaker popup
QA-674 TA: Fix - After deleting the project still showing name on homepage and setting icon, and clicking on setting no result showing.
QA-686 TA: Fix - Project Invited user is able to access the project setting by URL
QA-688 TA: Fix - Search user is not working on the Add User(s) to Project setting popup and its redirecting to the General Project Settings
QA-695 TA: Fix - Advanced search button stays there on the page even after search collapsed.
BE-1026 Fix - azure default voice amber doesn’t work
BE-1212 Fix - Telephony bot API does not support some symbols in Spanish
BE-1227 Fix - In telephony bot API, we should send START-INPUT-TIMER request only after the prompt is played fully.
BE-1229 Fix - In telephony bot, DTMF detection has very long delay
BE-1239 Fix - Audio server does not support prompt with email address
BE-1245 Fix - Bot logic does not get authToken in POST callback

Release 1.93.1

Changes:

BE-1207 Reject ASR requests if the default model doesn’t support languages specified in requests.

Release 1.93.0

New functionality in the Transcribe App:

BE-1100 TA: In large Video view put the Transcript (CC) in a frame that can be moved around the screen.
BE-1135 TA: Better copy-to-clipboard of the Meeting Minutes components
BE-1147 TA: Show “This Meeting has no video” in the large-video view if meeting has video
BE-1152 TA: In tag editor enable Save button only if something has changed

New functionality in other platform components:

BE-1032 Add GET /zoom/oauth and DELETE /zoom/oauth APIs
BE-1039 AIVR: Support DTMF in output actions
BE-1040 AIVR: Re-design for prompt playback - now using events instead on single audio stream
BE-1044 POST /auth-svc/device to return api URL in the QR code
BE-1068 Support misspellings in hints in REAL-TIME mode
BE-1159 Increase the number of allowed audio files in POST /asr/meeting to 25

Changes related to Integrity of Processing (fixes):

BE-1099 TA: Fix - Specified date format is not being used
BE-1110 TA: Modify NSIS Installer use proper setting of REINSTALLMODE
BE-1126 TA: Fix - Large-video chat popup captures focus and icons become not clickable
BE-1142 TA: Fix - When Zoom upload contains a video file we should set the “video” tag on this meeting
BE-1145 TA: Add beta label over the DOCX selector
BE-1146 TA: Fix - Remove UUID tool tips in the large-video view
BE-1162 Fix - Failure in sa-recompute on Edge
QA-545 TA: Now not be allowed to play multiple voice signature at the same time.
QA-567 TA: Fix - Files under processing status shouldn’t get option for (Re-Compute, Move or Delete)
QA-630 TA: Fix - Transcripts on Homepage are not updating after changing Projects.
QA-631 TA: Fix - Unable to edit Download permission after changing role from admin to user in one go.
BE-1166 In offline task, if we can’t download video, we generate audio-only dash-mpeg
QA-629 Console: Fix - Unable to download in JSON format in Download options in Transcribe under Transcribe+(beta)

Release 1.92.0

New functionality in the Transcribe App:

BE-779 TA: Show chat messages in Right-Hand pane
BE-856 TA: Fix - Login - not showing the locked until information
BE-883 TA: Zoom OAuth Handshake page added
BE-999 TA: Include video in the manual Zoom folder upload
BE-1003 TA: Implemented large video view
BE-1004 TA: New design for the Audio Source selector
BE-1010 TA: Add export in Docx format
BE-1014 TA: Better error massage in case of Recaptcha related Error
BE-1015 TA: Fixed the +tag message and added hover text
BE-1022 TA: Added a tag editor
BE-1053 TA: Edge: If SSO is enabled make login automatic if user hits /login url
BE-1055 TA: Show both installed version and available version of Zoom Meeting Assistant if newer version available
BE-1061 TA: Ability to Control Download permissions
BE-1063 TA: Ability to set a download permission on a User
BE-1082 TA: Meeting Chat now being shown
BE-1086 TA: Add Copy-to-Clipboard feature on the Overview page
BE-1087 TA: Add new download option for docx files
BE-1101 TA: Prevent entering regex that can match too much text
QA-488 TA: Make Meta Description Tags SEO friendly
QA-538 TA: Show message after user delete successfully

New functionality in other platform components:

BE-832 Implemented GET /asr/meeting/search/fields
BE-852 Digits formatter for real-time sessions
BE-862 Add new field AUDIO_SRC to meeting search API
BE-932 Use Google secret manager to manage credentials on GCP cloud
BE-974 In offline task, create gRPC channel to ml-svc on demand
BE-980 Ensure that Advanced Search queries only Projects/Contexts the User has access to
BE-988 App Selector: Add links to Privacy Policy and Terms of Use
BE-993 The video that is stored under videoId on a meeting now has audio removed
BE-1008 Console: Align Left and Right audio charts for the Telephony Bot Sessions
BE-1009 Console: Improve the look of the ASR settings forms
BE-1030 AIVR API: Add authToken to first Callback and use it in PUT /aivr/{ivrSid}/vars
BE-1031 AIVR API: Add ani parameter to GET /aivr/, add sorting, add endTime field
BE-1037 Configure MongoDB memory limit and cacheSizeGB
BE-1045 Add tags field to PUT /asr/meeting/{meetingId}
BE-1046 Modifications to GET and HEAD /data/{uuid}/file/{fnameWithExt} APIs
BE-1065 Enforce “download” permission in GET /asr/meeting/{meetingId}/transcript
BE-1069 Console: Show error if API to create new Edge Cluster fails
BE-1071 Support chat.msg in the response of GET /asr/meeting/uuid/data
BE-1079 Smarter match of chat speakers to the speakers in the Zoom Meeting (if there is no speaker timeline)
BE-1083 Add docx format option to GET /asr/meeting/{meetingId}/transcript
BE-1102 Prevent text redaction regex from matching too long patterns
BE-1103 Support smarter partial redacting PERSON if a name has no space

Changes related to Integrity of Processing (fixes):

BE-966 TA: Fix - Weird pause and play behavior on the Voice Signatures page
BE-967 TA: Fix - Missing Users step in new Project Wizard on Edge
BE-1078 TA: Fix - Password reset by admin does not work
BE-1080 TA: Fix - Unable to play audio/video in certain Edge deployments
QA-537 TA: Fix - Current Project is not picking correctly while move
BE-920 Fix - Meeting Search API exposes database structure in the error messages
BE-991 Fix - Meeting Search - Gt, Le Terms always returning empty results
BE-1012 Console: Fix Listen button from the Telephony Both Session view
BE-1034 Console: On Edge environments without HTTPS provide a workaround for copy to clipboard
BE-1038 Fix mongodb rolling deployment
BE-1056 Fix - Session gets stuck on certain corrupted audio files
BE-1076 Fixed: ascalon-cleanup fails to remove any orphan data object if persist=true is found in every data object in the first page
BE-1085 Fix - Search API shows meetings from projects that User has no access to
BE-1098 Fix - Modified meeting tags are not passed to the data in postgresql
QA-531 Console: Fix - On deleting the JWT success message is showing incorrect

Release 1.91.1

Changes:

BE-1013 Log invalid patterns found in RegexFormatters instead of throwing exceptions

Release 1.91.0

New functionality in the Transcribe App:

BE-71 Transcribe App: Show Meeting VIdeo
BE-709 Transcribe App: New meeting submitted by ZoomMA shows up on the home page automatically
BE-886 Transcribe App: Support new guides for Zoom Meeting Assistant Use
BE-961 Transcribe App: Remove Recompute button from Project settings - we can now select multiple Transcripts and recompute them
BE-975 Transcribe App: Save latest project opened by a user in clientSideProperties
BE-977 Transcribe App: Show Project info on the list of Shares
BE-986 Transcribe App: Re-enable API Security settings
BE-990 Transcribe App: Show Privacy Policy same way we show Terms of Service

New functionality in other platform components:

BE-739 Web Console: Log audit events
BE-778 In /ASR/meeting API add ability to upload meeting video and chat
BE-819 New Kappa Real-TIme model trained on IVR data
BE-901 Support more than one JWT per context
BE-925 Prepare fluentbit configuration for writing audit log to Grafana Loki
BE-945 Web Console: Show description for Edge Configurations
BE-946 Web Console: Show Model and Language information in the Context Dash
BE-956 Web Console: In Context Settings support multiple JWT
BE-987 Web Console: Change the text about available languages (now that we support Whisper)
BE-989 Web Console: Show privacy policy similar to how we show terms of service
BE-992 Handle MPD files in SegmentTemplate format
QA-490 Web Console (Edge): Launch Cloud Console in new tab
QA-516 App Selector: Update features list now that we have Video support

Changes related to Integrity of Processing (fixes):

BE-855 Web Console: Fixed transcript download in Transcribe+
BE-969 Web Console: Fix weird layout in Telephony App settings
BE-983 Fixed: REX tries to reserve Whisper model when diarization enabled
QA-310 Transcribe App: Fixed - Features and Usage are not being translated in selected language instead of English language
QA-467 Transcribe App: Fixed - Taking any URL as audio URL.
QA-505 Transcribe App: Fixed - Mouse hovering on the upload/ upload type always showing microphone recording.
QA-510 Transcribe App: Fixed - Editing any shared link changing its scope to public.
QA-511 Transcribe App: Fixed - After audio finish play icon should be paused.
QA-514 Transcribe App: Fixed - Hindi translation is on Recomputing the multiple transcription

Upload of video to the Transcribe App requires Voicegain Zoom Meeting Assistant version 1.2.1 or higher.

Release 1.90.1

Changes related to Integrity of Processing (fixes) All these apply to the Transcribe App:

QA-497 Fix: Accepting anything in tag, also taking space as tag.
BE-790 Fix: At the end of project wizard sometimes have to click Done twice to get the new project saved
BE-957 Fix: Not accepting MP4 file for transcription
BE-964 Fix: Calling Invoice API on Edge

Other changes in Transcribe App:

BE-955 Disable API Security Page

Release 1.90.0

New functionality in the Transcribe App:

BE-669 Transcribe App: Added audit log
BE-681 Transcribe App: Implement Person obfuscation in Meetings after a set time threshold
BE-824 Transcribe App: Add ability to select multiple transcripts from the table an perform selected Action on them
BE-843 Transcribe App: Add ability to specify Regex Redaction on the Project/Account Settings Text Redaction
BE-873 Transcribe App: Added multi-selector delete on the Shares list
BE-888 Transcribe App: Improve message on Zoom Meeting Assistant status
BE-894 Transcribe App: Change name of the exe file in installer bundle to install.exe
BE-909 Transcribe App: Make the digit formatting a default enabled option on creating new Project
BE-910 Transcribe App: Store collapsed status of home page elements in ClientSide Properties on a User
BE-915 Transcribe App: Add settings for Archival Text Redaction
BE-919 Transcribe App: Tweak behavior of the 4 selectors on Transcripts on Home Page
BE-929 Transcribe App: On Devices page - add a button to pair a Phone App

New functionality in other platform components:

BE-124 (EDGE) Make InfluxDB available for querying by customers
BE-778 In /ASR/meeting API add ability to upload meeting video and chat
BE-812 (SSO) Modify password reset page to match the style that is used in Web Console
BE-860 Support audio redaction also on DASH-MPEG audio
BE-872 Generate video dash-mpeg file in offline meeting task
BE-874 Support Accept-Ranges in the data presigned URL that returns mp3 audio
BE-877 Make sure the redacted audio for meeting sessions is mp3
BE-887 Modify POST /auth-svc/device to return QR code and also support new device type
BE-891 Add long-term redaction settings to the Account API
BE-892 Modify Meeting PERSON NER redaction to include speaker names
BE-893 Add internal meeting recompute API that runs recompute on the entire account
BE-904 Change how Context Regex and Account Regex are processed if both present
BE-905 Automatically enable digit formatting if any redaction is turned on
BE-912 Add fluentbit external helm chart to env-tracking
BE-916 Create a cronjob to call internal meeting recompute API on the entire account
BE-922 Update ingress and support basic-auth in influxdb helm chart
BE-933 OfflineMeetingWordsGrouper in ml-svc to use timestamp to group words if punc is missing
BE-937 Return raw transcription results if there is any error in formatter code
BE-942 Validate model and language in offline requests before submitting to offline queue
QA-458 Web Console: Add a button to delete a transcript under transcript beta.

Changes related to Integrity of Processing (fixes):

QA-469 Transcribe App: Fix multiple translation issues
QA-482 Transcribe App: Fix - Inactivity timout set on zero hrs and zero is showing out of the box, when toggle is disabled no value should show.
QA-497 Transcribe App: Fix for accepting anything in tag, also taking space as tag.
BE-870 Transcribe App: Fix - jumping to time in MP3 audio playback does not work
BE-890 Transcribe App: Fix for being logged out when accessing Speakers page with many voice signatures
BE-899 Transcribe App: Fix - Meeting audio redaction only works on the first 3 speakers in a meeting session
BE-935 Transcribe App: Fixed - incorrectly imposing share limit on Edge
BE-655 Fix for onprem-cluster-deployment task stuck when pods are in unexpected status
BE-735 Fixed: Recompute does not reapply PII Redaction
BE-895 Web Console: Fix - Regex redact options are saved but not displayed
BE-907 Admin Tool: Fix - When I suspend some account it is my account that gets suspended
BE-921 Fix: Dates with ‘the’/’of’ in the phrase don’t get classified as DMY NER
BE-928 Fix: Installer fails to apply settings to registry
BE-936 Fixed - List index out of range error when formatting certain Spanish results
BE-943 Fix weird output in case of multiple recompute with redaction and placeholder fill
BE-950 Fix bug in Spanish formatter : int() argument must be a string, a bytes-like object or a number, not ‘NoneType’

Release 1.89.2

Changes related to Integrity of Processing (fixes)

BE-882 TWIML sessions not closed if no TWIML Stop message sent on websocket

Release 1.89.1

Changes related to Integrity of Processing (fixes)

QA-477 Transcribe App: Sharing (public) a transcript and when try to play this transcript page refreshing all the time.
QA-483 Transcribe App: Not able to play the old transcripts

Additions:

BE-865 Add presigned url to response from GET /asr/meeting/shared/{shareId}/data

Release 1.89.0

New functionality in the Transcribe App:

BE-168 Transcribe App: Add ability to upload a directory with Zoom meeting recordings
BE-777 Transcribe App: Remove Project Transcripts page and show project transcripts on home page
BE-837 Transcribe App: Use GET /asr/meeting/search API on the home page
BE-845 Transcribe App: Show tags in full if just 1 or 2

New functionality in other platform components:

BE-785 Demo App: Now running on Firefox
BE-792 API: Allow audio data that is retrieved from url to use local settings similar to AuthConfig
BE-800 Return “lang” in Get Meeting Data API
BE-816 Web Console: Improve Contexts Card on the Context dash

Changes related to Integrity of Processing (fixes):

BE-794 Transcribe App: Fixed Word Cloud in Spanish
BE-799 Transcribe App: Fixed - summary and key items generated by recompute are always in English
BE-809 Transcribe App: Fixed - Enabling full masking (*****) caused Meeting Minutes to not be generated
BE-813 Transcribe App: Fixed - Wrong text shown if no Speakers shown on page
QA-418 Transcribe App: Fixed - While creating new project “Profanity Masking” tabs are showing hidden at 100% screen resolution
QA-426 Transcribe App: Fixed some missing German translations in the UI
QA-444 Transcribe App: Fixed - User detail is missing on the delete confirmation pop up, In case user delete the account from edit user popup
QA-448 Transcribe App: Fix - For multiple file uploads, no speakers show a difference even after selecting a range of speakers.
BE-774 Fix: Incorrect MPEG-DASH mpd format file generated
BE-786 Web Console: Fix color of buttons in Microphone Capture dialog
BE-798 Fixed: GET /asr/meeting/search assumes fromAllContexts is always true
BE-826 Web Console: Fixed - MRCP experiment page should be 100% wide
BE-827 Automatically delete REX pods in Unknown and UnexpectedAdmissionError status
BE-836 Fixed: /asr/meeting/search API should match all documents in textQuery parameter is missing
QA-423 Web Console: Fixed - Add Context window closing when trying an add context with a name that already exist.
QA-432 Web Console: Fixed in MRCP ASR - Create Grammar is not working and also the file format is not defined
QA-433 Web Console: Fixed in MRCP ASR - On the experiment How to Upload video is not playing

Release 1.88.1

Known Issues:

BE-809 Enabling full masking (*****) caused Meeting Minutes to not be generated - If you enable Full Text Redaction for some category then Meeting Minutes may not be generated. A work-around is to disable Word Cloud on the Project.

Changes related to Integrity of Processing (fixes)

BE-774 Fix: Incorrect MPEG-DASH mpd format file generated
BE-802 (TA) Fix: Account redacting formatters are saved to Projects
BE-803 Fix: Formatters defined on account are not being used in POST /asr/meeting
BE-806 (TA) Fix: When generating a pdf, NullPointerException is thrown if meetingMinutes is not enabled
QA-405 Fix: Demo app has incorrect recaptcha key

Release 1.88.0

Backwards incompatibility : This release removes sentences parameter from session.content in /asr/transcribe API. It gets replaces with segments.

New functionality in the Transcribe App:

BE-416 (TA) Login from Transcribe App to Freshdesk
BE-640 (TA) Support renaming existing Speaker
BE-685 (TA) Admin (Account level) controlled PII redaction
BE-699 (TA) more options for PII Redaction (add full and partial masking)
BE-703 (TA) Implement https://transcribe.voicegain.ai/freshdesk/login
BE-704 (TA) Change the URL for Help->Submit a Support Ticket
BE-721 (TA) Add PII Redaction page inside Account settings
BE-732 (TA) Add x to cancel a search for transcript within project
BE-751 (TA) Support upload of multiple files for transcription
BE-752 (TA) Add two filter toggles on the Speakers list
BE-755 (TA) On home page show selected Project together with Home
BE-756 (TA) Allow change of name of an External Speaker
BE-757 (TA) Show email for User Speaker
BE-759 (TA) Modify the list of entities to redact (ZIP, DMY)
BE-764 (TA) Show only current project on the LH menu
BE-793 (TA) 3 distinct URLs for “Submit Support Ticket” link depending on environment
QA-369 (TA) Improved Error message if user tries to sign-up second time with same email
QA-373 (TA) Improved title of page with Account Users
QA-400 (TA) Disable Save button if no new users have been selected

New functionality in other platform components:

BE-666 Return information about what parts of the transcript were redacted and by what rule
BE-688 Better end-of-segment detection for real-time sentence-by-sentence (segment-by-segment) mode
BE-707 Enable keepalive ping also on the websocket used for the audio
BE-710 Add formatters field to the Account API
BE-728 Presigned URLs for Data Object files
BE-729 In the API remove “sentences” as content and replace with new “segments”
BE-738 Partial redaction of PERSON NER -> generate Initials
BE-749 Add defaultContext to User data in User API
BE-769 Make GET /data/{uuid/file/{fnameWIthExt) return special output if contentType is application/dash+xml

Changes related to Integrity of Processing (fixes):

BE-708 (TA) Fix: Cannot delete a device that is shown in the device list
BE-715 (TA) Fix: At larger zoom levels Left-Hand menu unusable
BE-761 (TA) Fix: Project Transcript list shows meeting from home page after auto refresh due to STOP message
BE-791 (TA) Fix: We are not generating Meeting Minutes in German if the Project is German
BE-795 (TA) Fix: Weird long topics for Spanish Meetings
QA-365 (TA) Fix: Next Audio button not responding on Single Click (or with delay)
QA-368 (TA) Fix: Search and Project Name is overlapping over each other
QA-370 (TA) Fix: Trying to recompute any file, resulting all files showing across all projects.
QA-372 (TA) Fix: Transcripts table-Table sorting by “Name” and “Created on” is not working.
QA-374 (TA) Fix: My shares table-Table sorting not working for name, transcription date, and scope.
QA-377 (TA) Fix: Search text should reset on Project switch
QA-401 (TA) Fix: Other users table- Sorting by last active is not working.
QA-403 (TA) Fix: other accounts users table- Sorting by name is not working.
QA-410 (TA) Fix: Left menu is breaking down when selecting it rapidly.
QA-417 (TA) Fix: Account Text redaction/project setting- When placeholder is selected fields should be required fields.
BE-720 Fix issue with UUID on Oracle MongoDB
BE-723 Fix: ex-autoscaler cronjob is not working when there are more than 9 rex instances
BE-742 Fix: asr-api sends hypothesis msg of the next sentence to websocket before the recognition msg of the current sentence
BE-746 Fix: Some words are included in two recognition results
BE-750 Fix: In some case, asr-api does not send recognition websocket msg when getting EOS
BE-753 Logic fix in EZInit kubelet configuration
BE-754 Fix: audio redaction is not working in meeting API
BE-762 Fix: Redaction.originalValue is returned with default debug level
BE-763 Fix: Prevent storing duplicate formatters
BE-797 Fix: Incorrect decimal formatting if currency
QA-363 (Web Console) Fix: New User Wizard step4
QA-379 (Web Console) Fix: In Experiments sections, input fields overlapping with close icon
QA-380 (Web Console) Fix: IVR Proxy status is not showing properly
QA-392 (Web Console) Fix: Properly notify about attempt to create duplicate name Context
QA-398 (Web Console) Fix: textboxes are not cleaned up when I try to create a second context

Release 1.87.1

Changes related to Integrity of Processing (fixes)

BE-722 Recognizer does not send RECOGNIZE-COMPLETE if no audio has been sent to Recognizer

Release 1.87.0

New functionality in the Transcribe App:

BE-603 Transcribe App: Automatically refresh status of the transcripts
BE-621 Transcribe App: Hide invalid transcript move project destinations
BE-623 Transcribe App: Improved Error message in case of transcript move error
BE-626 Transcribe App: Limit Basic Plan user to only 5 shares
BE-628 Transcribe App: Enhanced Project selection
BE-638 Transcribe App: PII Redaction placeholder validators allow non-ASCII characters
BE-651 Transcribe App: Add Table of Contents in the exported PDF
BE-657 Transcribe App: Add Metadata in the exported PDF
BE-668 Transcribe App: Zoom Meeting Assistant Download name now has version included
BE-671 Transcribe App: Add ability to completely disable Download on Edge
BE-675 Transcribe App: Update content of Home Page if there are no transcripts
BE-680 Transcribe App: Add ability to completely disable transcript copy to clipboard on Edge
BE-693 Transcribe App: Update content of Home Page if there are no transcripts
BE-694 Transcribe App: Add Login and SignUp buttons to the Share Expired/Invalid page
BE-695 Transcribe App: Do not show the Move option for transcripts in a project that does not belong to a user and has not been shared with the user

New functionality in other platform components:

BE-574 Web Developer Console upgrade to latest AntD
BE-641 Disabled deprecated support for the Language Models in the Developer Console
BE-667 Admin Tool: Allow modification of Rate-Limits
BE-678 Easy scaling for real-time transcription supported
BE-687 Add noAudioTimeout option to asr API

Changes related to Integrity of Processing (fixes):

BE-620 Transcribe App fix: Sorting shares by Expiry time
BE-635 Fixed issues in Walk-Through Wizard in the Developer Console
BE-637 Transcribe App fix: Prevent invalid PII Redaction placeholders in Project settings during creation
BE-639 Fix access to a transcript from a Developer Console Context Dash
BE-672 Transcribe App fix: Prev/Next buttons not working correctly
BE-674 Fix missing audio charts in Transcribe+ in Web Console
BE-691 GET /user/sync API checks user existence based on user.email
QA-261 Transcribe App fix: Make transcript not clickable if status is Queued
QA-357 Fix weird behavior of Web Console if the current context is deleted
QA-362 Transcribe App fix: Sorting by transcript Status

Release 1.86.5

This release enables higher number of replicas for autoscaling.

It is just a configuration change.

Release 1.86.4

Changes related to Integrity of Processing (fixes)

BE-662 Temporary fix for “ASR-API kills REX session if there is no activity for 5 minutes” - timeout extended to 15 minutes
BE-663 Fix for: asr-api on CHD environment rejects derived session because callback url is redis
- this was introduced by: BE-615 Enable TWILIO protocol in CHD environment (sapi.voicegain.ai)

Release 1.86.3

Changes related to Integrity of Processing (fixes)

BE-631 Remove punctuation between two digits in En BERT punctuation model

Release 1.86.2

This release includes an improved real-time English model delivering higher accuracy of speech recognition.

Changes related to Integrity of Processing (fixes)

BE-622 (SSO) Fix issue with Spinner that does not stop if login failed (no error message was shown)

Release 1.86.1

This release has several improvements in the PDF export in the Taanscribe App (Meeting API)

Changes related to Integrity of Processing (fixes)

QA-301 Transcribe App: Removed recompute option from transcripts in Error state.
QA-308 Transcribe App: Fixed getting blank screen on My Shares
QA-323 Transcribe App: Fixed project flag indicators not working as expected on Homepage (also BE-611)
QA-324 Transcribe App: Fixed unable to move transcript between projects (also BE-616)
BE-612 Transcribe App: Incorrect area outlines on the Home Page

Fixes in other components:

BE-614 Web Console: In Edge Config prevent selection of configuration not matching the version
BE-615 Enable TWILIO protocol in CHD environment (sapi.voicegain.ai)

Release 1.86.0

New functionality in the Transcribe App:

BE-550 Transcribe App: Usability improvement in share creation.
BE-458 Transcribe App: Add a checkmark to indicate among the matching Speakers the one that is assigned to the Speaker from transcript
BE-494 Transcribe App: Added validation for values entered in the text redaction fields
BE-555 Added support for multiple languages to the UI of the Transcribe App.
BE-568 Transcribe App: Delete all devices belonging to a User when a User is being deleted
BE-570 Transcribe App: Improvements to the projects view in the Left-Hand menu
BE-571 Transcribe App: Add new Zoom Icon (with status) in left-hand Menu
BE-576 Transcribe App: Add name of the user to be deleted to the User Deletion confirmation dialog
BE-582 Transcribe App: Modified text on the Signup page for EDGE if OIDC SSO is enabled on the account
BE-596 Transcribe: Add support for Hindi language for transcription.
BE-605 Transcribe App: Improved instructions on ZoomMA download page
BE-598 Transcribe App: Added PDF export option (beta)
BE-569 Transcribe App: Change the Home Page Plan info to a new format
BE-575 Transcribe App: Replace Browser Capture Icon with Google Meet Icon
BE-580 Transcribe App: Move Latest News to a separate page that in accessible from the Help Menu

New functionality in other platform components:

BE-561 Admin tool: Added ability to change the account Billing Style
BE-563 Added API to add credit to an account
BE-558 Implemented a static method that generates PDF from a meeting JSON
BE-564 Make it possible to modify billing style on the Account using Admin tool
BE-565 Added pdf format to GET /asr/meeting/{meetingId}/transcript API
BE-546 Support stopping billing-utility from processing anything
BE-547 new-billing-utility to store usage of each TranscribApp account in Firestore
BE-554 new-billing-utility supports auto-refill

Changes related to Integrity of Processing (fixes):

BE-440 Web Console: fix missing waveform in the microphone capture preview
BE-591 Make sure no account information is revealed by password reset API.
BE-514 Added Matomo to App Selector with correct IDs for dev and prod
BE-567 GET /user/{userId} returns 404 if the specified user does not exist
BE-581 prevent new-billing-utility from processing multiple hourly storage requests
BE-560 EZ Script fixes for K8s 1.27 and deprecated config
BE-572 App Selector: Update text on Signup page
BE-587 Reject with 400 (Bad Request) all requests to sapi endpoint which have audio.capture=true
BE-590 Accept WSS streaming protocol on PCI/CHD environment

Release 1.85.2

Changes related to Integrity of Processing (fixes)

BE-549 Public Transcribe App share from EDGE cycles non-stop if used without login

Release 1.85.1

Changes related to Integrity of Processing (fixes)

BE-548 Do not send multipart/form-data parameters as files in audio.callback. Only audio data should be sent as file.

Release 1.85.0

New functionality in this release in the Transcribe App:

BE-45 Add Sharing of transcripts from Transcribe App
BE-422 In Transcribe App add Custom behavior for Help->Submit a Support Ticket option
BE-423 In Transcribe App: Support sharing from the Transcript Detail Page
BE-424 In Transcribe App: Add Page where the user can see all the meetings that they have shared
BE-426 In Transcribe App added Page that shows a single shared transcript to a not logged-in user
BE-427 In Transcribe App added Page that shows a single shared transcript to a logged-in user
BE-449 In Transcribe App: add ability to recompute Meeting Minutes for all Meeting withing Project

New functionality in other platform components:

BE-225 Modify the Admin Tool to show “Last Used” time
BE-233 In Admin tool: add PCI-DSS column to the main account table display
BE-326 In Web Console: Show SHA sum of the Edge Cluster configuration
BE-387 In Web Console: If account is PCI compliant then show CHD Rate Limits
BE-417 Implemented Share APIs
BE-418 Implemented GET api for meeting data that uses shareId
BE-425 Add sharedBy and sharedWith parameters to GET /asr/meeting query
BE-447 Add PUT /asr/meeting/context/{contextId}/recompute method to recompute all the meetings on a context
BE-474 In Admin Tool: Show the number of Phones
BE-475 In Admin Tool: Add sorting by balance
BE-476 In Admin Tool: On the Usage chart make it possible to deselect one or more lines
BE-484 In Admin Tool: Add a “view password” feature to the login dialog
BE-485 In Admin Tool: Add “Suspend Selected” account functionality
BE-487 In Web Console Signup: Add confirmation of the values being submitted
BE-489 When creating new buckets enable versioning and set lifecycle rules
BE-514 Add Matomo to App Selector
BE-515 In App Selector: Add support for selectable Spanish/English language in the UI
BE-528 In Admin Tool: Add filtering on null value for Last Login and Last Used fields
BE-539 Support a new property: minio.server.region
BE-541 Add Region to Object Store settings for Edge Cluster

Changes related to Integrity of Processing (fixes):

QA-183 In Transcribe App fix the link from “Manage Users”
BE-197 Add Kubeadm API update to EZUpdate.py
BE-411 In Web Console: Add splash loading screen
BE-446 In Transcribe App improve prompt in the URL entry box
BE-453 In Transcribe App fixed the “missing voice signature speaker” error
BE-454 Configured brotli for Cloud ACP and Transcribe App
BE-456 Fixed new-billing-utility fail to store an instance of Account in Redis due to oidcSettings
BE-462 Fixed new-billing-utility fail to send alerts due to a missing environment variable
BE-463 Fixed new-billing-utility fail to process storage and phone number usage periodically
BE-466 Fixed multiple records of storage usage of an account are written to postgres
BE-467 new-billing-utility submits the usage of a phone number to Billing System
BE-477 In Web Console in Add User dialog fixed the User option showing twice
BE-478 new-billing-utility fails to group the same type of ASR usage into one single record for a given interval
BE-479 new-billing-utility derives billing_to_process from only asr usages and ignores other types of usages
BE-480 Fixed: Account query API does not return correct values for pciDss field
BE-483 In Admin Tool: Better names for tabs
BE-492 new-billing-utility to ensure each quantity has up to 6 decimal places before it’s sent to Fusebill
BE-493 In Web Console: fixed issue with New User Wizard (Joyride)
BE-505 In Admin Tool: Make login safer by using pre-login feature
BE-512 GET /config-cluster fails to include clusterConfigShaSum in the response
BE-517 Fixed: EZINit issues with home directory and autossh for auto generated voicegain user.
BE-518 rex fails to invoke SessionMeasurementUtility.commit() for some realtime ASR sessions
BE-521 Fixed in Web Console: Transcribe+ table does not span full available width
BE-525 Fixed Polling URL returned from POST /sa
BE-527 Fix in Transcribe App: Download no longer downloads TXT and Audio
BE-529 In Admin Tool: Change filtering values on Contexts, Edge, Websockets, Phone #, and Users columns
BE-537 In Web Console: Fix user gets duplicated when adding roles
BE-538 Fixed in Transcribe App: the request for meetings on the home page should retrieve only 20 meetings
BE-540 Fixed in EDGE Web Console: Fallback Login is shown incorrectly
BE-542 Fix RedisTimeoutException: Unable to acquire connection!
BE-544 Fixed: Cloud function cannot add task to redis

Third-Party vulnerability (security) related fixes:

BE-464 Admin Tool: Update to yarn and update packages to newer versions
VM-79 Unimrcp docker vulnerability
VM-81 freeswitch docker image vulnerability
VM-133 Python package protobuf==3.20.1
VM-135 offline-main docker image vulnerability (ffmpeg)
VM-136 ml-svc-grpc docker image vulnerability
VM-149 (SSO) nth-check Regular Expression Denial of Service (ReDoS)
VM-150 (SSO) async-validator Regular Expression Denial of Service (ReDoS)
VM-151 (SSO) word-wrap Regular Expression Denial of Service (ReDoS)
VM-152 (SSO) css-what Regular Expression Denial of Service (ReDoS)
VM-153 filebeat docker image vulnerability
VM-154 elasticsearch-master docker image vulnerability
VM-155 Triton docker image vulnerability
VM-156 Java services Docker image vulnerability
VM-157 httpd docker image httpd:2.4.57-alpine3.17 vulnerability
VM-158 redocly/redoc:v2.0.0-rc.50 docker image vulnerability
VM-159 transmogrifier ssh and web docker container vulnerability
VM-160 ingress-nginx==4.6.0 helm chart vulnerability
VM-161 freeswitch 2.3.10 docker image vulnerability
VM-162 python:3.9.16-slim@sha256:78740d6c888f2e6cb466760b3373eb35bb3f1059cf1f1b5ab0fbec9a0331a03d docker image vulnerability
VM-163 nats:2.9.16-debian-11-r1 docker vulnerability
VM-164 telegraf:1.26.1 docker image vulnerability
VM-165 mysql:5.7.30 docker image vulnerability
VM-166 Pod container allows privilege escalation on exec
VM-167 Pod container is allowed to run as root
VM-168 grafana 6.6.0 docker image vulnerability
VM-169 single-tensorflow-serving:2.11.1 Docker image vulnerability

Release 1.84.3

Changes related to Integrity of Processing (fixes):

Fix several UI glitches that were introduced when updating AntD version.

Changes related to Security:

BE-220 demo.voicegain.ai requires Content Security Policy and related Headers

Release 1.84.2

Changes related to Integrity of Processing (fixes)

BE-448 Project transcript view shows also Home Page transcripts directly after switch from Home to Project view

Release 1.84.1

Changes related to Integrity of Processing (fixes)

BE-442 Fixed in Transcribe App: Voice Signature sample playback always plays from the beginning of audio
BE-443 Fixed in Transcribe App: Transcription fails because unable to find the speaker

Release 1.84.0

This minor release includes a significantly more accurate Spanish real-time model.

Here is a list of fixes and changes:

BE-224 Track on each account the most recent access
BE-242 Finalized the new style of the Browser Capture pop-up in Transcribe App
BE-260 In Web Console: In File Transcribe Dialog add ability to enter hint misspellings and boost
BE-279 Allow only Audio files for upload to Clip Manager
BE-281 In Transcribe App: Added a page with Terms of Use that is reachable via the Help menu
BE-288 Fixed in Transcribe App: Asks to setup a payment method while payment method already setup on the account
BE-327 Show context id in the card with JWT token
BE-336 Improved English NER
BE-365 Voicegain SSO App updated to latest 3rd-party package versions, also switched from npm to yarn
BE-367 Make the User Management table sortable in the Web Console
BE-369 Generate JWT with userId and use it where needed
BE-370 Fixed in meeting minutes: Summaries truncated mid sentence
BE-371 Fixed: Save button not enabled on EDGE in user profile
BE-372 In Transcribe App left-hand menu - prevent overlapping scroll bars
BE-373 fixed in Transcribe App: Error when deleting last Device
BE-374 Fix in Transcribe App: asr-api doesn’t allow users to move a meeting session to a different context in some case
BE-377 Library Vulnerability check for the SSO app
BE-378 Modify EZInit and EZUpdate script to correctly set directory permissions
BE-379 Modify /asr/transcribe/async API to support audio calbacks
BE-380 Add report of the audioCallback in the full /asr/transcribe/async results
BE-381 Add DMY to list of NERs that we support for redaction in asr/transcribe formatter
BE-382 Fixed in Transcribe App: incorrect speaker names found in exported meeting transcription
BE-384 Supports audioCallback with PI Redacted audio
BE-388 Fix in Web Console: Transcription from URL request is missing audio.capture parameter
BE-389 Improved Context selection dialog in the Web Console
BE-391 In Transcribe App: trim spaces from entered URL
BE-392 Disable Password Recovery Key page if user is logged in using SSO
BE-394 New API to return AuthConfig by name from Context
BE-396 Do not offer local password change to users logged in via SSO
BE-401 In Web Console: add ‘S3 Compatible’ type of auth config
BE-404 Support context.description in Context API
BE-407 Changes to authConfig to support S3 Compatible URI
BE-408 Fixed in Transcribe App: Wrong confirmation dialog show when deleting user from Account
BE-412 When syncing users, avoid copying a user to edge if the user’s email is found on edge
BE-414 API for multipart/form-data for Audio Callback
BE-415 Add GET /asr/transcribe/status/queue API
BE-419 Add customValues to the OnPrem Cluster API
BE-421 Fixed bug: In regex-based redaction only the first regex is being used
BE-429 Fixed formatting anomaly: ok, let’s try number 1
BE-431 Fixed in Transcribe App: Bad URL in Advisor
BE-435 In Web Console: When we are done creating new context - switch to the new context
BE-436 Fixed in Web Console: broken validation for area-code numbers
BE-439 In Web Console: fix formatting of the Transcribe Detail page

Release 1.83.0

Key things that are new in the Transcribe App this minor release:

Meeting summaries are generated by GPT-3.5
Zoom Meeting Assistant now uses Device Pairing to authenticate with the Transcribe App. You will need to download and install Zoom Meeting Assistant ver 1.0.0+
On EDGE, Transcribe App now supports OIDC SSO.

Here is a list of fixes and changes

BE-164 Modifying Edge configuration is no longer supported - on change new configuration needs to be created
BE-211 In Web Console - do not accept characters not needed in field values
BE-228 Fixed Major 3rd-party vulnerabilities
BE-241 OIDC SSO functionality added to Transcribe App
BE-246 Page to configure the OIDC SSO parameters in Transcribe App
BE-247 New Login page for Transcribe App if SSO is enabled
BE-252 Zoom Meeting assistant ver. 1.0.0 with new Device Pairing functionality
BE-264 Invite User dialog now extended to handle SSO users
BE-266 New welcome to Transcribe App email for the SSO invite case
BE-267 If SSO is enabled for the Transcribe App then Forgot Password should direct to SSO
BE-273 Meeting transcription will complete even in presence of NATS messaging errors
BE-275 APIs to support device pairing
BE-282 Transcribe App now support device pairing - a new page for device management
BE-283 New Zoom Meeting App installer now supports properties needed for device pairing
BE-291 Fixed bug: Auth Configuration may overwrite good credential with a bad one
BE-292 Meeting Minutes summary now generated by LLM
BE-298 New Auth API methods to support SSO login
BE-305 New field in /cluster API to support Edge deployments with SSO
BE-311 CHD environment can submit measurements to local influxDB
BE-312 New Zoom Meeting Assistant installer supported on Edge
BE-316 Track CHD sessions for billing
BE-319 Rate-Limit measurements are now sent from CHD environment and tagged with chd=true
BE-324 Add CHD boolean tag to all measurements written to the influxDb
BE-325 New Rate Limits in the Account API for CHD environment
BE-330 Separate Rate-Limits for CHD environment
BE-332 Apply fetchTimeout to the ffmpeg -timeout option
BE-338 Automatically signup users from SSO if the email domain is allowed
BE-339 Support cases where an SSO user wants to login but the domain is not allowed
BE-340 In Transcribe App: support key sentences which are summaries as cannot be attributed to a single speaker.
BE-341 In Transcribe App Meeting Minutes: implement speaker substitution also for key sentences
BE-342 Fixed TTS (Google voices) fails on text like e.g. “7, 6, 2, 6, 2”
BE-345 If OIDC is enabled for a Cluster then do not show “Signup Code” in from Account Settings
BE-346 In Transcribe App, provide a pop-up alert if a device need pairing approval
BE-347 GET /confgroup now accepts JWT tokens
BE-348 Changes to how authConfig is updated using PUT/confgroup/uuid
BE-349 In POST /asr/meeting: do not override context to null if JWT has no context
BE-353 fixed: PUT /confgroup/uuid with languages=[null] causes NPE
BE-356 Modify PUT /user/uuid to accept JWT tokens for authentication
BE-357 Unable to delete values of “allow signup from domains”
BE-359 Support DMY NER in Transcribe App project configuration
BE-360 In Transcribe App: New user invited from SSO will set name auto populated

Release 1.82.4

This release allows for use of the /confgroup API with JWT token.

Release 1.82.3

This release upgrades Kubernetes to 1.25

Release 1.82.2

This maintenance release has the following changes:

BE-256 Modified App Selector to properly render on mobile browsers
BE-257 Added notice to Login Screen on Transcribe App on mobile browsers

Release 1.82.1

This maintenance release has the following changes:

BE-243 Fix for not getting any transcriptions from the Telephony Bot sessions
BE-244 Fix for erratic recognition in the Telephony Bot

This release also fixes several 3rd-party vulnerabilities.

Release 1.82.0

This minor release chas the following changes:

BE-73 Add info about language to Mic Recording and File Upload dialogs
BE-74 Improved Project Move dialog
BE-76 Improved Zoom Meeting Assistant download page
BE-80 Improved/Simplified Browser Capture Preview in Transcribe App
BE-81 New Project Wizard in Transcribe App upon first login
BE-109 Fixed: Issue with saving meeting with certain commas in the name
BE-110 Option to send transcription results back over the audio websocket
BE-113 EZInit Script disable Leader Election on NFS Provisioner
BE-114 EZInit updated for Containerd
BE-115 Upgrade minio to highest backwards-compatible version
BE-119 Improved the throughput for Kappa model when using web-socket to stream audio
BE-130 Clarified that only Screen and Browser Tab audio capture works
BE-133 AccountUsage generation complete in new-billing-utility
BE-135 New Billing Utility writes to postgresql the same data that would be written to Fusebill
BE-138 Fixed old portal URL still referenced from within Grafana
BE-143 Fixed incorrect ERROR tooltip visible for new uploaded Voice Signatures
BE-146 Fix concurrency issue in mrcp_rex plugin history
BE-148 The Zoom App download page is now correctly reporting the App version
BE-150 Disabled Automatic creation of a New Transcribe App project upon first login
BE-151 every GET data API request decreases rate limit by 1 (instead of 10)
BE-153 Google Storage bucket creation now works in the QA environment
BE-154 Remove confusing Logs option from LH menu in Edge mode
BE-155 In Web Console on EDGE: Remove all Log options from the LH menu
BE-162 Fix Terms of Service it highlight on the LH menu
BE-163 Offer First Project Wizard when user logs in and has no Projects
BE-165 Updated Content-Security-Policies for QA Environment
BE-170 Usability improvements to App Selector (reduce mouse clicks)
BE-172 Add ability to turn off the current language reminders in Mic and Upload dialogs
BE-174 Ensure that returned items from the Extended Summary (topics) are unique
BE-175 Improve reliability of Browser Capture
BE-178 Show tooltip with Creator name when hovering over avatar
BE-182 In Transcribe App: Remove diarization from the Browser Capture preview
BE-184 Better location for the close-search-box icon
BE-188 In Web Console: Make email on the Phone Management Page clickable link
BE-193 Fixed: Unable to go from Apps page direct to Project Transcripts page
BE-194 Support Statefulset uniMRCP deployment (for Cloud)
BE-200 Fixed upload of corpus files for Language Model
BE-202 Fixed uploading an audio file on demo.voicegain.ai
BE-207 In Demo App: Allow more audio file types for Upload
BE-209 Add back button to Pricing Plans page in Transcribe App
BE-212 In Transcribe App: show creator name tooltip when hovering over creator avatar
BE-213 Made email in balance details clickable (mailto:)
BE-214 In Demo App: Remove FFW and REV buttons
BE-219 In Web Console: Move password reset link to the first step
BE-227 Added a link to submit a support ticket in Transcribe App
BE-229 Added new field “complianceType” to Account

Release 1.81.0

This minor release has the following fixes and improvements in the Transcribe App:

BE-21 Fix expanded left-hand menu obscuring page content if frozen
BE-24 In user Profile add a setting to hide language flags on Project icons
BE-26 New Project design for the Transcribe App (mainly left-hand menu)
BE-27 Add Search within transcript in the detail view in Transcribe App
BE-52 Flags on Projects are inconsistent
BE-59 Back button doesn’t work
BE-63 Transcribe from URL not working in Transcribe App
BE-64 Upgrade on home page of the Transcribe App takes you to Password Reset
BE-65 Plan Upgrade page working off stale billing data
BE-66 Walk-Through Wizard cannot be launched
BE-68 Hide expanded LH menu upon a click
BE-112 None or not all Projects are visible on Edge
BE-126 Clicking on the Downgrade button has no effect
BE-127 Show a Tool Tip for Project Settings
BE-128 Sending password confirmation in clear-text
BE-140 Modify behavior of the expandable LH menu - do not expand when I merely move the mouse across it
BE-141 Show tooltip when hovering over project icon in the transcript list on home page

This release has the following fixes and improvements in the rest of the Voicegain Platform:

BE-48 Issue with AccountUsage collector query
BE-53 Issue with AccountUsage collector query (2)
BE-56 Duplicate/Triplicate Projects visible in Transcribe App
BE-62 Text formatter occasionally throws an Exception
BE-117 Increase the limit for the number of concurrent websockets
BE-118 Remove duplicate Transcribe App projects on Edge
BE-129 Reduce real-time delay in case of many concurrent sessions
BE-146 Fix concurrency issue in mrcp_rex plugin history

Release 1.80.2

This maintenance release has the following fixes:

BE-106 SMU is using a lot of Redis memory

Release 1.80.1

This maintenance release has the following fixes

BE-51 Fix for WS/WSS protocol for Audio Streaming is not working
BE-54 Fix session_duration set to 0 for each offline ASR session measurement in influxDB
BE-58 Add missing insecure websocket url to Ingress

Release 1.80.0

This minor release has back-end improvements that will eventually give better monitoring more precise billing/usage data. It also includes a model that is more accurate on call-center audio. We also made changes to support new Billing Plans in the Transcribe App.

BE-1 Fixed number formatting in multi-speaker transcripts
BE-2 Fixed error connecting to Redisson caused by 3rd-party lib
BE-3 Fix issues where in some cases mixed meeting audio was missing some channels
BE-6 Fix measurements not written to influxDB due to field format mismatch
BE-7 Fix measurements not written to influxDB due to field format mismatch
BE-9 Fix issue where measurements with identical timestamp would overwrite
BE-15 Resolve issue with field type being inconsistent with old data
BE-30 Fix service startup problem due to invalid character in authorization value
BE-39 Configure production-ready log levels on classes that write to influxDB
BE-17 Configure Cloud influxDB connection parameters for all environments
BE-18 Configure influxDB writing classes correctly for deployment on Edge
BE-19 MeasurementWriter to be able to distinguish if it is running in Cloud or on Edge
BE-20 Add diarization and channel_selector tags to api_session_asr measurements
BE-33 AccountUsage generator should use longer query interval if run for the very first time
BE-36 Deploy latest offline model VoiceGain-omega-x - trained on additional call-center audio
BE-38 Changed pricing and time limits on Transcribe App plans

Release 1.79.0

This minor release has several back-end improvements to the core Voicegain Platform:

The rate limits use is now logged to the influxDB and can be queried from Grafana. You can configure alerts in Grafana so that you can know if you are getting close to the rate limits.
The throughput in the offline mode has been further improved. Edge deployments have been tested to 1200 hours transcribed per hour for extended periods of time.
Cause for the spike in redis use under extreme loads has been identified and removed. This will result in better stability due to much lower use of the redis resource.
Edge deployments to GCP VPC now require smaller set of permissions to use Google Storage.
In Edge deployment, MongoDB indexes are created on startup. The set of indexes used has been optimized.
Removed permessage-deflate support on websocket connections used by real-time transcription. This in order to reduce the latency.

Release 1.78.1

This maintenance release includes the following for the Transcribe App:

Keyword and phrase highlighting in the transcript text
Bug #671 fixed: home/login page loop in all the nonces somehow are removed from the indexedDB

Release 1.78.0

In this minor release the throughput of the offline transcription in two-channel mode has been significantly improved.

This release addresses the following issue for the rest of the Voicegain platform:

#82: (ACP) Add matomo to Web Console (dev and prod)
#86: (CMP) the Detail option takes me to the Login page instead of the Account Detail page
#212: ascalon-asr-api cannot start on CHD environment
#214: Get error when running offline transcribe on CHD
#216: Add code that logs the measurements for usage of rate-limits per account
#217: Log usage of the offlineQueueSizeLimit
#220: Edge deployments to GCP VPC now support Google Storage via S3 adapter
#223: Add language to the result of Transcription
#225: Random text output from real-time transcription in some cases after long silence

Release 1.77.1

This maintenance release includes:

Back-end update of Redis to 6.2.10 which has a fix to for the issue https://github.com/redis/redis/pull/11590
More accurate real-time model

Release 1.77.0

This release adds Zoom Meeting Assistant status display to the Transcribe App. Due to this change this release will require Zoom Meeting Assistant version 0.3.0 at minimum.

Transcribe App issues addressed:

#555: Automatically insert underscores into multi-word hints
#667: Do not hide microphone and browser capture icons for old projects that have no language setting value

Admin tool has been improved in this release:

Last Login Time and Billing Type columns have been added.

This release addresses the following issue for the rest of the Voicegain platform:

#211: nats-websocket fail to send messages to topic subscribers.

Release 1.76.0

This minor release has the improved Spanish Real-Time model.

Transcribe App issues addressed:

#406: Add ability to delete a keyword in Speech Analytics settings
#461: The downloaded audio file is mp3, but the extension is wav
#532: In languages Project settings show if microphone transcription is not available for given language
#610: Change the lock-out message to show the local time instead of the UTC.
#665: Fix weird behavior in Settings if old Project has no Meeting Minutes enabled
#666: Add a hide button for the Zoom Meeting Assistant banner

Release 1.75.2

This maintenance release includes:

Fix for Issue #208 which was affecting Signup via the Developer Console
Improved demo.voicegain.ai app with improved security

Release 1.75.1

This maintenance release fixes the following Transcribe App issue:

#204: Occasional Error in Recomputing status

Release 1.75.0

This minor release has the improved Spanish Offline model.

This release adds ability to Recompute meetings in the Transcribe App and in the /asr/meeting API. This means that all the NLU processing of the transcript can be redone after, e.g., change in the Project analytics settings.

Other changes to the Transcribe App:

Info about available Zoom Meeting Assistant is shown on the home page.
List of search results shows also topics for each transcript.
Removed Update button on the Key Items settings. Key Items will always be updated to the latest configuration.
Hid Key Items configuration from non-Admin users.

Transcribe App issues fixed:

#647: If a user has no signatures and I hover over the playback icon for a signature it shows “Error” tool tip
#656: When creating a new Project select a color by default - absence of a default selection was a bit confusing.
#664: Logout is not working - this was only in case if the app ran out of nonces.

Other changes include:

Improved Admin Tool - better able to handle the thousands of accounts that we have.
Fully tested dash-mpeg support. Will be coming soon to the Transcribe App.

This release addresses the following issues for the rest of the Voicegain platform:

#198: Suboptimal http code returned in case of bad password reset
#199: asr-api returns 401 for POST /asr/meeting/async, which is unsupported
#200: Upgrade Redisson version due to a bug
#201: When GET /sa/config is called we should update the key items to most recent version
#203: New API method to recompute a meeting: /asr/meeting/{meetingId}/recompute

Release 1.74.1

This maintenance release fixes the following issues:

offline task fails on re-try due to format value type mismatch
some sessions in ERROR state not removed from session cache as soon as ERROR determined

Release 1.74.0

This Minor Release enables User self-signup via email on Transcribe App on Edge.

More options available in the OFFLINE /asr/transcribe API settings.fomatters:

enhanced - provides formatting for entities like: URL, EMAIL, PHONE, SSN , CC
redact - provides redaction for entities like: EMAIL, PHONE, SSN , CC, PERSON, ZIP
regex - text redaction/modification using regular expression matching

EZUpdate script has been added for updating the Ubuntu packages on the Edge servers.

This release addresses the following issues in the Voicegain platform:

#182: for sapi.voicegain.ai - limit what is accepted as values of audio.source
#183: for sapi.voicegain.ai - do not accept the session.portal parameter
#184: for sapi.voicegain.ai - reduced range for session.poll.persist
#185: for sapi.voicegain.ai - reduced set of parameters for session.websocket
#190: GET /public/monitor/asr returns HTTP 406
#191: GET /account/uuid fails to return submissions and saDemoContext

Release 1.73.1

This is a dummy maintenance Release, meant to test the automated release/deployment pipeline changes.

Release 1.73.0

This minor release enables Browser Capture in the Transcribe App on Edge.

This release also addresses the following Transcribe App issues:

#075: Pause is behaving like mute
#625: Add “No Microphone” option in browser Capture
#627: (Meeting Minutes) Modifications to the KeyItems settings in Project
#631: (Meeting Minutes) Support negative examples and regex
#633: Cancel out of the Mic Save dialog does not work
#634: (Meeting Minutes) Add Enabled toggle for each Key Item

This release addresses the following issues for the rest of the Voicegain platform:

#107: Wrong content-type in response from https://console.voicegain.ai/ascalon-web-api/public/monitor/asr
#139: If account has allowSignupsFromDomain values then allow creation of new users only with valid emails
#144: Return simple account data from GET /account/{uuid} if no authentication provided
#156: Make freeswitch-esl-client one of our internal libraries
#162: Add Cache-Control: no-store to all API GET responses that may change
#163: Implement POST /asr/meeting/search
#166: data-api to avoid including sosRef in the response of GET /data/uuid if its audio file is missing
#169: Add mpdId to /asr/meeting API
#170: Additions to formatters in the API
#172: New pciDss field on account
#173: Option to generate PCI-DSS restricted JWT token
#174: Spelling error in JWT aud field
#176: Add negativeExample and negativeRegex to saConfig
#177: Secure GET /account
#179: Add disabled field to saConfig Phrase
#186: Add rejection justification for meeting key items
#188: (Edge) negativeExample value not saved correctly
Diarization inference reliability has been improved.

Release 1.72.0

This minor release offers improved Key Items feature within the Meeting Minutes in the Transcribe App.

This release also addresses the following Transcribe App issues:

#593: Subscribe to notifications from back-end and log the incoming messages
#615: (Edge) make some Profile fields read-only on Edge
#619: Add a button for applying defaults to KeyItems
#620: Add a check for validity of regex to the Key Items Example dialog
#621: Switch from regex to advancedRegex for key items
#622: Show projects alphabetically sorted

This release addresses the following issues for the rest of the Voicegain platform:

#145: Add releaseVersionDetail to response from GET /spec
#146: Add PUT /sa/config/{saConfigId}/defaults
#147: Return Phrases in SA Config in fixed order
#148: Validate regex loaded from the built-in Key-Items JSON

Release 1.71.1

This maintenance release fixes 3 issues:

#143: Weird clock values when on Edge and logged in as a user with role User
#617: (Edge) New created project not visible in the project list
#618: (Edge) New created project shows files from an older project

With this release we also support a minimal, self-contained, docker-compose deployment of MRCP ASR on GPUs.

Release 1.71.0

This minor release adds configurability of the Key Items of the Meeting Minutes in the Transcribe App.

This release also addresses the following Transcribe App issues:

#556: Prepopulate the Name with the file name
#599: Show project name on transcript detail view
#602: Show avatar if Speaker is a User
#605: In Overview view, clicking on the times in the Summary should scroll to relevant Section
#607: Remove the front-end bolding of keywords in key sentences (done in back-end now more accurately)

This release addresses the following issues for the rest of the Voicegain platform:

#128: Add justification fields in meetingMinutes
#129: Support json-mc in GET /asr/transcribe/{sessionId}/transcript
#130: Speed up GET /asr/meeting
#131: Add metatada to PUT /asr/meeting
#132: Prepopulate SA Config with Meeting Minutes phrases
#133: Meeting Minutes defaults for existing projects
#136: Change the lock-out duration for user sign-in

Other changes in this release are:

Updated EZ-Init script
CallHome License Server

Release 1.70.2

This maintenance release fixes one issue:

#126: Add project filtering to meeting search for normal Users

Release 1.70.1

This maintenance release has only two changes:

Diarization processing has been moved to GPU which is of interest to Edge users because it will reduce the CPU resource requirements.
In Transcribe App, the default setting of Meeting Minutes on existing Projects has been changed to make it less confusing.

Release 1.70.0

This minor release addresses the following Transcribe App issues:

#595: Highlight the topics/keywords in the key sentences in Meeting Minutes
#596: Add audio playback and transcript text links to the sentences in Overview and Key Items
#600: In Profile: Separate Account from User settings
#601: In New Project Wizard set the toggle for Meeting Minutes to enabled by default
#603: (Edge only) disable Download if transcribeAppSettings has disableDownload:true

This release addresses the following issues for the rest of the Voicegain platform:

#103: CC number marking in logs
#111: sttGenericData in AudioCodes API now supports maxAlternatives
#112: sttGenericData in AudioCodes API now supports interpretation
#123: Added POST /data/audio/ws method to data API
#124: Use Content-Type: text/plain; charset=utf-8 for all text/plain responses
#125: Added transcribeAppSettings to OnPrem Cluster

Moreover the following improvements were done in the ML back-end.

Punctuation generation now runs on GPU instead of CPU

Release 1.69.0

This minor release adds two major new features in the Transcribe App:

Meeting Minutes - an AI generated overview of the meeting including key sections with topics/keywords and key sentences in 4 categories: Actions, issues, Risks, Requirements. Meeting Minutes can be enabled from Project Settings.
Text search of transcripts - from the Home page find transcripts containing the specified words.

Other Transcribe App issues addressed in this release:

#530: Fix confusing text in the confirmation dialog for user deletion
#560: Fixed a scroll issue where the last line of transript was only partially visible.
#578: Do not allow Phrase example sensitivity values outside the 0-1 range
#580: Fixed a glitch in multi-column formatting
#584: Made user email not editable - the User dialog incorrectly was allowing it to be modified
#585: Added scroll shortcuts to the right pane in the Transcript Detail view
#588: (Edge only) Added Sync from Cloud button which can be used to sync all the user info from Cloud to the Edge.
#589: (Edge only) Added Send Test Email button. This is in preparation of the rollout of the SMTP email support on Edge.

This release addresses the following issues for the rest of the Voicegain platform:

#104: New setting for transcription and recognition APIs that controls logging of results
#117: Add localOnly parameter to GET /model/acoustic

Release 1.68.1

This maintenance release supports multiple languages choice in the /asr/transcribe/async and /asr/recognize/async. It runs recognition using all specified languages and returns the transcript with the highest confidence.

Release 1.68.0

The key features in this minor release are:

Improved diarization in the text areas where one speaker switches to another.
More accurate speaker Voice Signature matching.

This release fixes numerous small Web UI issues, mainly usability, in the Transcribe App related to the Voice Signature functionality made first available in the 2 previous releases.

The installer for the Zoom Meeting Assistant is now signed with Extended Validation Certificate.

This release addresses the following issues for the rest of the Voicegain platform:

#97: When deleting a Speaker we should remove references to it from Context voiceSignatureSpeakers
#98: Orphan cleanup delete the audio clip for speaker signature

Release 1.67.0

This minor release addresses these specific Transcribe App issues:

#546: Older projects do not generate topics
#547: Saving a project switches it to a different one
#550: Assign transcript speaker to a known Speaker (with or without voice signature) - also create new Speaker if needed
#551: Add page for managing Speakers

The release addresses these specific Speech Analaytics App issues:

#12: Turn off pagination in the Calls page in Speech Analytics App - this is temporary measure before switching to new pagination style
#14: Upload file should use short-lived JWT

This release addresses the following issues for the rest of the Voicegain platform:

#81: Enhance logging for AIVR sessions
#86: When creating a new User let’s create a corresponding Speaker entry (for Transcribe App accounts)
#88: Add support for vsConf and originalName (meeting API)
#91: Add mustHaveSignature parameter to GET /speakers API

Release 1.66.1

This maintenance release includes improved punctuation for the English Offline transcription.

Release 1.66.0

This minor release provides an improved model for Offline English transcription - you will notice about 1% higher accuracy

The release provides these overall improvements to the Transcribe App:

Voice Signature support. Users can now create voice signatures and be automatically recognized in transcripts. Next release will add support for creating Voice Signatures for other speakers based on audio.
Phrase detection in transcript is now suported.
Topic detection has been improved. Still a beta feature.

The release addresses these specific Transcribe App issues:

#514: use /auth/login/pre to get the key to encrypt password for login
#519: tweak space around arrows pointing to other speakers text
#521: do not show the expiry info if the expiry time is “Never”
#540: browser capture - same text in both channels
#542: sometimes getting a blank page when opening transcript (Cannot read properties of undefined (reading ‘name’))

The current version of the Zoom Meeting Assistant going with this release is 0.2.17

This release addresses the following issues for the rest of the Voicegain platform:

#69: Changes to pagination of API results (/calls API)
#70: add mine and sharedWithMe fields to context
#71: fail to persist AbstractGrammar.type in MongoDB
#76: Add new style pagination to GET /asr/meeting
#77: support PII Redaction in Meeting API
#78: new auth API: /auth/prk/filename
#79: add format parameter to GET /sa/calls (to support CSV export)
#80: MeetingSession.sizeInStorage includes the size of each original audio file by mistake

Release 1.65.0

This minor release addresses the following Transcribe App issues:

#364: Add option to move transcript to a different project
#466: run GET /data/{data-object-id} before submitting microphone recording for offline - this ensures that mic recording is available
#505-506: better tooltips regarding Project
#508: different behavior when user selects project while on Home page
#509: on all of the Project Settings pages show the Project name the same way we do it on e.g. Transcripts page
#510-511: (Edge only) add ability to reset a password using Password Recovery Key (PRK)
#513: Front-end should allow apostrophe in the Project name
#515: change the “data not found” text to “No transcripts”
#516: Changes on the Apps Download page

This release addresses the following Developer Web Console issues:

#74: Add info about Rate Limits on the Account
#78: Password is encrypted even on HTTP connections (relevant to Edge only)

This release addresses the following issues for the rest of the Voicegain platform:

#63: new /speaker API
#64: small changes to /asr/meeting API to take account new Speaker objects
#66: add voiceSignatureSpeakers to POST /asr/meeting
#67: add voiceSignatureSpeakers to context APIs
#68: Allow DataObject to be associated with an Account

Release 1.64.1

This maintenance release includes an improved model for diarization which will provide better diarization accuracy.

The release also fixes these issues in the ML Services:

#1: Words that should be stop words appear in the Topic results
#2: Topic extraction should be case insensitive.

Release 1.64.0

This release addresses the following Transcribe App issues:

#469: Added options for profanity masking and digits formatting
#488: “My Transcripts” Project has no Creator value set
#487: Add owner information on Project Settings page
#489: On the Cloud, for Basic Account, do not show User setting for a Project
#490: Creator avatars vanish when I do search
#491 through #498: All these make it easier to identify current Project and create a distinction between a Home page and a Project Transcript List page
#499: Show non-speaking participants on the Transcript Details page - Requires Zoom Meeting Assistant ver 0.2.11 or higher
#501: First project for a new user should not be called “My Transcripts” but “{firstname}’s Transcripts”

Zoom Meeting Assistant installer is now digitally signed.

This release addresses the following Developer Web Console issues:

#76: New Auth Configuration type - support for Twilio encrypted recordings

This release addresses the following issues for the rest of the Voicegain platform:

#58: New field in authConfig on Context - support for Twilio encrypted recordings
#59: Add support for authConf in POST and PUT /data/audio (twilio-rec-encrypted type)
#60: NPE in REAL-TIME transcribe session is no audio has been received at all over websocket

Release 1.63.0

This minor release addresses the following Transcribe App issues:

#382: (re-opened) wrong (previous) audio being played
#403: reduce spacing in the Keyword dialog
#404: Tags are not displayed
#447: add a password view icon to Transcribe App login
#456: App download page does not show on Chrome on Mac
#462: in Project settings - if switching the project, the settings do not switch to reflect the selected project
#468: Increase max file size for upload to 256MB
#473: change of terminology when removing user from Project
#476: Lag in switching between audio being played
#477: Add Role setting to Edit User dialog
#478: Add to user Profile a toggle that makes the account discoverable
#480: change header: Latest Activity => Latest activity across all Projects
#483: Allow someone with role=User to share their own Projects with other users
#484: handle a case where words are empty
#485: change wording on the transcript delete form

Note: This release requires Zoom Meeting Assistant version 0.2.3 or higher.

This release addresses the following Developer Web Console issues:

#72: modify file upload to use JWT in the POST /data/file API
#72: change refill amounts

This release addresses the following issues for the rest of the Voicegain platform:

#47: Increase audio file size for upload to 256MB
#52: New discoverable field on User
#53: APIs to support for Password Recovery Key
#54: data-api rejects a request when a context doesn’t have any allowedOrigins
#56: add minRequiredZoomMAVersion to /spec API

Release 1.62.1

This maintenance release fixes the following issue:

#rcj-555: fail to copy asr transcription result files to a new context - affects SA App Demos

Release 1.62.0

This minor release addresses the following Transcribe App issues:

#380: only positive (>0) values of speakers should be allowed
#398: on Zoom download page add info about needed Zoom settings
#407: add ability to delete a Keyword example
#424: (Edge) Cannot see the projects I created
#436: A wrong value is assigned to user.memberOf
#440: I am an Admin user and in Projects settings I do not see the icon for Users
#449: Project->Settings->Users page show a “no users in project” page briefly before showing that there are users
#450: deleted project still visible in the project menu
#451: American English is highlighted instead of Spanish which is the actual setting on the Context
#452: (Edge) Not able to save changes to file name
#453: Show error message as a tooltip over Error
#454: Cannot select other languages on Cloud Transcribe App
#457: (Edge) New URL to the Knowledge Base for Transcribe App on Edge
#459: selection of a column for sorting a table cannot be undone
#460: sorting by storage size is as if storage size was a string not a number
#463: back button does not take you back
#464: on projects settings - move NER to bottom of page
#467: modify logic for checking validity of Labels
#469: Allow max number of speakers for diarization to be 12 instead of 10
#472: (Edge) The latest sessions are not on the home page
#474: no way to add another user to project

Known issue in Trancribe App:

Uploading files larger than 128MB may fail in the Web Browser.

This release addresses the following issues for the rest of the Voicegain platform:

#39: number formatting is not working on meeting transcripts
#40: Some sessions are on the second column when there is only one speaker
#41: fail to read formatters from MongoDB
#42: create a new project if user logs in from=TranscribeApp and there are no private contexts with type=Transcription for that user
#46: queue size rate-limit being exceeded
#47: Increase audio file size for upload to 256MB
#48: increase maxspeakers for diarization to 12
#49: add nonSpeakingParticipants to /asr/meeting API
#50: change in allowed format for Labels
#51: new GET /spec API

Release 1.61.0

This minor release includes a major overhaul of the Transcribe App:

Transcribe App now supports Zoom Meeting Assistant App which can submit all your local Zoom Meeting recordings for transcription to the Transcribe App.
New Transcript Detail display that no longer requires paging and supports transcript from overlapping speech.
Transcribe App is now available on Edge - this way you can deploy it to all users in an Enterprise and keep all your confidential data local.

Other minor changes include:

Decrease playout of the Telephony Bot prompts. Latency of actions following a prompt has been reduced by about 800ms.
Built-in number grammar has a fix for a minor issue.

This release includes improved offline speech-recognition model with about 1.5% improvement in accuracy.

Release 1.60.4

This maintenance release changes 2 things relevant for MRCP users:

The content on NLSML returned in MRCP results - We now return <nomatch/> element within <result><interpretation> in case of Completion-Cause: 001 no-match. In the past <nomatch/> was not included, instead an NLSML corresponding to a normal recognition was returned even in case of Completion-Cause: 001 no-match
GRXML grammars with scope=private attribute now parse without an error. Note, however, that the private scope is still not enforced. That will be fixed in future releases.

Release 1.60.3

This maintenance release fixes:

#rcj-546: DataObjects not being deleted by cleanup task. Fixes the de-referencing issue.

Release 1.60.2

This maintenance release includes:

Usage data now stores session Tags, so if you request from us, e.g., a monthly usage report then each session in the report will have those tags.

Issues fixed:

#rcj-545: re-register listener whenever FirestoreException is received in FirestoreWebApiConfigCollectionEventListener – The problem manifested itself in new JWT tokens not fully working for offline-sessions - the requests were accepted but final result was Error. The cause was Google Firestore client occasionally getting an error event from Firestore upon which no subsequent events would pass through. We implemented a workaround where the listener re-registers in case of error.

Release 1.60.1

This maintenance release includes:

In Transcribe App:
- Small speed-up of the transcript loading
- Tool-tips for prev/next buttons
In Web Developer Console:
- Significantly speeds up loading of long Transcripts and Meetings
- Private/Shared context functionality has been improved.
- Removal of the password change reminder from the Edge Web Console - this was to support Edge deployments which are not connected to Internet.

Release 1.60.0

This minor release includes:

Support for sttSpeechContexts in AudioCodes API - can be used to pass hints to recognition.
- Ability to pass additional parameters through sttGenericData parameter coming soon
Password strength check compliant with latest NIST guidelines.

Release 1.59.2

This maintenance release includes:

Improved NLU for IVR prompt intent detection (contact support@voicegain.ai if you are interested in this feature)

Issues fixed:

#rcj-538: NPE in AudioCodes API when websocket closed
#tsw-3: Default language selection in new Transcribe App project

Release 1.59.1

This maintenance release includes:

Ability to configure external document and object storage for Edge deployment
Maximum persist time for meeting transcripts on Cloud has been increased from 7 to 31 days (this is in /asr/meeting API)
Added password view toggle in the Web Console Login
Multiple upgrades to 3rd-party dependencies to remove known vulnerabilities

Release 1.59.0

Issues fixed in this minor release:

#358: (Transcribe App) Login not working if email is not lower case
#ocp-780: Topic generation for SA not working

This release also changes the logging library used by our back-end code.

Release 1.58.1

This maintenance release adds in Transcribe App:

Ability to set the transcript expiry to “Never expire”
Improved pop-up preview for transcript in browser-capture mode.

Release 1.58.0

This minor release includes:

Even more accurate offline transcription
New “rho” model which gives accurate, low latency real-time transcription (meant for API use, enabled using "acousticModelRealTime" : "VoiceGain-rho").
Custom profanity list feature - please contact us if the default profanity list does not work for your use case.
Expired-item cleanup code - starting from June invoice you may see charges for storage - previously they were not included because cleanup code was not 100% tested.

Issues fixed:

#rcj-536: Stop users from logging into a different application from account.type
#rcj-534: handling unknown speaker it transcript export

Release 1.57.0

This minor release includes:

More accurate offline transcription.
Improved language selection in Transcribe App.
Private Contexts in Web Console

Issues fixed:

#rcj-532: Microphone capture transcript in Transcribe App occasionally does not load (No Data)
#rcj-527: Text export from Transcribe App does not have modified speaker names
#rcw-6: (Transcribe app) An exception shows up in the console when Settings is clicked

Release 1.56.0

This minor release includes:

More accurate offline/batch model. Accuracy is improved in particular in lecture/zoom meeting/podcast type of audio.
Hints in large-vocabulary continuous recognition via MRCP.
For MRCP customers using large-vocabulary continuous recognition - we now can deliver custom intent recognition using NLU.

Issues fixed:

Transcribe App
- #346: Word Cloud is now shown even though the data is there
- #347: Named entities (NER) are not highlighted correctly

Release 1.55.1

This maintenance release fixes the following issue:

#ocp-779: offline process returns NOINPUT for audio shorter than 1.25s

Release 1.55.0

This minor release includes:

Web Console now supports Meeting view - for viewing transcripts from /asr/meeting API
Alpha version of the Zoom Recorder utility is available upon requests - it captures individual audio of each Zoom Meeting participant and submits it for transcription to /asr/meeting API. It works both with Voicegain Cloud and Edge.
Improvement to internal ml-svc request routing - previously in certain scenarios the requests to ml-svc service might end up going to just a subset of available pods thus reducing performance.

Issues addressed:

#rcj-522: GET /asr/transcribe/{uuid}/transcript fails intermittently
#rcj-520: support replacing acoustic model(s) with languages
- this is in place to accommodate older pre-1.53.0 style requests which reference names of acoustic models that no longer exists (have been merged into model bundles)
- new style requests should use languages parameter which is sufficient in most cases

Release 1.54.1

This maintenance release partially addresses the following issue:

#rcj-517: old/expired data cleanup inefficiency

Release 1.54.0

This minor release includes:

Edge deployment improvements:
- Support for external S3-compatible object storage
- Support for SSL Certificates
- Support for Kubernetes clusters with no GPU
- Configurable Grafana Dashboard for visualizing API use
Other improvements:
- Beta version of new /asr/meeting API suitable for transcribing per-speaker audio from e.g. a Zoom Meeting
- Beta version of encryption for the /data API
- Faster offline transcribe, including optimized pipeline for audio coming from S3 (or other external URL)
- Improved English acoustic model (offline mode) - about 1.5% better accuracy on meeting / lecture type of audio
- Beta version of a German offline acoustic model
- Language settings in the Web Console improved - only acoustic models compatible with a selected language are shown
- Added profanity masking option to the Transcribe Dialog

Backwards incompatibility:

the reuse parameter in the /data API will now be ignored - each POST request will create a new Data Object

Fixed issues:

#ocp-777: When afterlife is 0, offline process does not submit ERROR status to asr-api (no error callback)
#rcj-484: better error handling for Fusebill
#rcj-496: audio redaction no longer works for mono audio

Release 1.53.3

This maintenance release adds:

improved food-kappa model

This release fixes the following issue:

#rcj-488: spanish offline transcription ignores languages defined in the associated context

Release 1.53.2

This release adds:

Choice of 4 high quality TTS voices in Indian English (en-IN)

This release fixes the following issue:

#rcj-486: support using context.asrSettingsRecognition.acousticModelRealTime for Rex resource reservation for AIVR sessions if present

Release 1.53.1

This maintenance release addresses the following issue:

#TranscribeApp-338: Fix countries images icon at Transcribe App - Acoustic model selection

Release 1.53.0

This minor release includes:

Telephony Bot API uses a more efficient method to stream audio to the recognizer. The response latency has been reduced a bit.
Model selection mechanism has been changed. A single Acoustic Model can now support multiple languages; therefore, we added a extra language parameter that can be provided together with a model name.
MRCP ASR now supports a dual language mode - it is possible to do recognition not knowing what language is spoken. Currently, only en/es combination is available.
Profanity masking has been added to the Transcribe API
In Web Console - the audio display now shows milliseconds in tooltip and when zoomed in.

Release 1.52.0

This minor release includes:

Browser Capture feature available in the transcribe app. This allows you to capture full audio of e.g. Zoom or MS Teams meetings.

This release addresses the following issues:

#rcj-477: AIVR - attempt to play empty questionPrompt results in error (was re-opened)
#rcj-475: Unable to listen to recordings of Telephone Bot API sessions

Release 1.51.0

This minor release provides the following more accurate models:

English real-time (streaming)
Spanish off-line

This release addresses the following issues:

#rcj-477: AIVR - attempt to play empty questionPrompt results in error
#rcj-476: AIVR -the last prompt in disconnect is not being played and the hangup is not done
#rcj-474: REX returns incorrect long result for certain utterances in recent benchmark
#vgp-833: The startTime field datatype is “String in the spec, but web-api returns integer
#rcj-472: Get exception when submit silence to SA
#rcj-470: Cannot GET SA transcript
#vgp-830: DEL SA session and DEL SA config need to be in the public API spec
#ocp-773: offline-process occasionally return empty result
#vgp-831: “name” field should be required when creating SA config
#rcj-469: attempt to get /sa results returns INTERNAL_SERVER_ERROR
#ocp-770: diarization is not returned when audio is short

Two enhancements:

Old and/or orphan data cleanup has been moved to separate task and made faster
Time filter has been added to the GET /data query method

Release 1.50.1

This maintenance release fixes the following issue:

#rcj-463: call duration reported from offline transcription to billing is 1/1000th of what it should be

Release 1.50.0

This minor release addresses the following issues and enhancements:

#rcj-461: the body of a callback request with callback.format=text is empty
#rcj-460: better handling of billing system rate-limiting errors
#rcj-459: new audio selector for call center calls to be processed offline

Release 1.49.3

This maintenance release fixes the following issues:

#rcj-457: Missing influxDB entries due to identical timestamp
#rcj-456: Long audio files (e.g. 24 hours) cause OOM
- Max audio file duration has been limited to 8 hours
#rcj-455: Offline session time reported to billing as real-time session

Release 1.49.2

This maintenance release modified how offline transcription billing information is submitted to Billing in order to:

make it more efficient,
correct underbilling.

Release 1.49.1

This maintenance release upgrades the Redis server and clients to a newer version and makes changes to their configuration to prevent certain issues affecting High Availability.

Release 1.49.0

This minor release has numerous back-end changes to improve Voicegain platform ability to handle large loads.

This release fixes the following issues:

#rcj-448: User logs not available.

Release 1.48.4

This maintenance release adds the following features:

vadMode “disabled” is now supported - this turns off the VAD and may give higher accuracy where there is constant speech in presence of high volume background noise/hum
offline processor configuration has been modified to handle more offline requests

This release includes the latest offline model. For info about improved accuracy see update on this blog post.

Release 1.48.3

This maintenance release fixes the following issues:

#rcj-447: New Language Model build not working due to Cloud Function issue/change
#ocp-768: offline process cannot transcribe mlp and truehd file because of rate limiting bug
- This applies also to other formats which do not provide audio duration information. In addition to these files not being transcribed, other files were delayed in processing.
#rcj-445: NullPointerException while processing sync /asr/recognize and /asr/transcribe requests with dataStore as audio

Release 1.48.2

This maintenance release fixes the following issue:

#rcj-444: Unnecessary check for audio file format

Release 1.48.1

This maintenance release fixes the following issues (which impacted Edge deployments with no Internet connection):

#rcj-442: Consider empty context model name same as null
#rcj-441: Web Console goes blank if unable to obtain status info

Release 1.48.0

This minor release adds or changes:

Punctuation/capitalization and digit formatting for Spanish Transcription is now available.
Rate limit for Edge offline queue throughput is now expressed per hour rather than per day.
Modifications were made to allow for transcription on Edge without connection to Internet.

This release fixes the following issues:

#rcj-439: responses of polling requests don’t include word.spk for diarized requests
#rcj-435: NullPointerException when polling a transcription session

Release 1.47.1

This maintenance release fixes the following issues:

#rcj-434: web-api throws LowAccountBalanceException in an Edge environment with a port-based license
#rcj-433: endTimeOfCurrentBillingPeriod is found in some Developer-only accounts in prod and dev environments
#rcj-432: NoHttpResponseException: some callbacks failing - old http library
#rcj-431: ConverterNotFoundException: No converter found capable of converting from type [java.lang.Integer] to type [java.time.Instant]
#rcj-430: An offline SA session (mono, no diarization) is not showing up in Dev Console
#rcj-428: ACH payments not working

Release 1.47.0

This minor release adds:

Ability to host more than one model per GPU - this is of importance to Edge users who no longer will have to allocate one GPU per model.
Improved transcript paragraph splits both in the Web Console and in the downloaded TXT file
Improved accuracy of the Spanish offline model (the real-time model is still available only upon request)
Improved backwards compatibility of the Cloud with the Edge deployments - it mainly relates to login behavior (SSO, etc).
Improved EZInit script for Edge installs. Two core improvements are:
- uses a voicegain user for deployment - in the past it was using one of existing user accounts on the server
- has been adapted and tested with Ubuntu Server - although we still recommend using Ubuntu Desktop due to its better GPU support

Backwards incompatibility:

requests to /asr/transcribe/async API with portal field value but no portal.label specified will fail - portal.label is now a required value

This release fixes the following issues:

#rcj-427: web-api should reject requests with invalid dataobject UUID
#ocp-767: offline process throughput rate-limit checking error when ffmpeg cannot detect input audio duration

Release 1.46.1

This maintenance release fixes the following issues:

#rcj-419: POST /sa sessions failing (stuck in PROCESSING)
#rcj-418: account login API returns fields incompatible with older Edge versions
#rcj-417: first callback always fails after idle

Release 1.46.0

This minor release starts to enforce rate limiting. For details of about rate limiting, including the default values, please see this knowledge base article.

This release adds:

ability to disable punctuation and capitalization in the output (rcj-411)
session specific diarization settings for transcription (rcj-398)
longPersist parameter to POST /data API (rcj-397)
Second real-time model (faster but less accurate): VoiceGain-rho-en-us

This release fixes the following issues:

rcj-414: Remove archaic words from British English dictionary
rcj-396: Polling request is slow when there are multiple web-api instances

Release 1.45.3

This maintenance release includes the following changes:

Results of speech recognition which get assigned the __garbage__ semantic tag will now be returned as NOMATCH irrespective of the confidence value w.r.t threshold.
Diarization has been made more memory efficient
API documentation includes info about rate limiting

Release 1.45.2

This maintenance release fixes the following issues:

#rcj-393: New session takes few extra msec to start after a longer pause between request
#rcj-379: customer reports that Expired transcripts are still visible (Edge deployment)

Release 1.45.1

This maintenance release fixes the following issues:

#rcj-386: multiple callback requests are sent for a given request
#rcj-385: occasionally responses from recognition take extra long (incompleteTimeout) - incompleteTimeout was not behaving as per spec
#rcj-384: Realtime session takes 1-2s to start - requests to billing were not being cached efficiently
#rcj-383: Audio from AIVR transcription shows in the Portal under normal Transcribe

Release 1.45.0

The release has these main changes:

Introduces a uniform way to handle kubeconfg in Edge deployments. This makes it easy to deploy Voicegain Edge to various Cloud Platforms.
For Edge: moved to a different GPU runtime framework which uses GPU resource more efficiently. Allows to run more recognition sessions on same hardware.
Optimized offline task queue - significantly higher throughput is now possible (more hours of audio transcribed in the same period of time).
The latency of the callback response in real-time transcription has been reduced to better support voicebot scenarios.
Polling now goes via load-balancer URL instead of directly to individual services in order to better support rolling deployments.
Microphone capture in web browser applications moved away from deprecated methods

The release fixes the following issues:

#rcj-368: callback from recognition not working (introduced when adding redis:// callback method)
#rcj-365: Transcripts of Telephony Bot Session cannot be viewed in Web Console
#rcj-359: Reduce delay between end of recognition and call-back response

Known issue that is not fixed in this release:

#rcj-380: polling requests fail to return transcript after poll.afterlife – the workaround is to make sure that content.full includes both [“transcript”, “words”]
#rcj-377: if content.full is only [“transcript”] then the callback will not have the transcript – it is necessary to specify [“transcript”, “words”] to get any transcript in the callback

It also provides these enhancements:

#vgp-822: naming changes in Edge management ACP pages - uses less ambiguous names in Web Console

Additional announcements:

We now offer advanced monitoring for Edge deployments (for a slight per-port price premium).

Release 1.44.1

This maintenance release provides:

Further improved accuracy of the real-time/command model.

Release 1.44.0

This minor release provides:

New versions of the models used for real-time and offline transcription. They provide improved accuracy. This improvement in accuracy is particularly noticeable for short real-time transcripts, like e.g. encountered in voice-bots.
The hints feature provides new misspellings capability. For more info about hints see this Knowledge Base article

This release fixes the following issues:

#rcj-358: web-api returns “missing websocket name” by mistake
#rcj-355: ContentType.word-tree was serialized as WordTree by mistake

Release 1.43.6

This maintenance release fixes the following issue:

#vgp-821: availableFeatureDetails field not retrieved properly by Edge Web Console

Release 1.43.5

This maintenance release fixes the following issues:

#rcj-353: completeTimeout not working in SEMI-REAL-TIME mode
#rcj-351: submitting an offline SA session with stereo audio ended up with mono audio - this bug was introduced in release 1.42.0
#vgp-820: allow : in hint names - new weight property

Release 1.43.4

This maintenance release fixes the following issues:

Certain fields not saved on a transcription record due to NPE
Uploaded LM corpus files incorrectly reported as having wrong format/encoding
Edge Web Console not showing all Application Modes correctly

Release 1.43.3

This release fixes issues:

#rcj-349: Acoustic model setting in context is not used in offline transcribe

For Transcribe App released fixes for the following:

Issue with navigating away from microphone capture which would break the recording
Not able to upgrade in a single step from monthly to annual and to a higher Plan
Computation of remaining days till usage reset
Minor issues in password entry for a new password
Several small UI issues

Release 1.43.2

This release fixes issues:

#rcj-345: On ACP, Edge configuration selection is not filtered by selected version
#rcj-347: Invalid value for persist, must be a value less than or equal to 604800000

For Transcribe App:

Fixed several small UI issues
Fixed pricing values shown on Billing Plans page - was reporting price/user/month as price/month
Fixed double counting usage for microphone transcriptions.
Fixed error when doing transcription with Expiry set to longer than 1 Week.

Release 1.43.1

This maintenance release:

Changes signup text for Developer Web Console from “300 free monthly minutes available” to “$50 one-time credit available”

Release 1.43.0

This minor release:

Significantly improves Hint and Language Model functionality. You can now achieve much higher accuracy using Hints and/or Language Model
Available configurations for Edge deployments now are annotated with the compatible release range - the Web UI ensures that you can apply to your Edge cluster only compatible configurations
New accounts get $50 credit instead of the monthly 300 minutes allowance. Old accounts retain 300 minutes allowance.

Two new Applications are being released:

Speech Analytics App - analyze call-center calls
Transcribe App - transcribe audio from meetings and other audio files

See www.voicegain.ai for how to signup.

Issues fixed:

Memory issue affecting Offline transcription which occasionally would cause the transcription to be stuck in Processing

Release 1.42.1

This maintenance release:

Reduces the amount of memory used by the offline transcription tasks leaving more space that may be used for transcoding audio files. Should reduce chance of failed offline transcription.

Edge specific notes:

It is recommended to apply this release to your Edge deployment if you are doing offline-transcription.
This release is not relevant to Edge users using only MRCP ASR or only real-time transcription.

Release 1.42.0

This minor release adds the following:

OFFLINE transcription now uses ffmpeg to transcode audio - see the list of supported audio formats
Edge Deployment UI now shows all available versions and allows deployment of any one of them onto an Edge Cluster. Version rollback is also possible using this feature.
data-api has been split off from the web-api - this improves robustness of the platform
Digit formatter has been added to transcript returned from polling real-time transcription session
In the Cloud, added ability to increase data persistence time for individual accounts. In the past persistence was capped at 7 days.

Edge specific notes:

data-api has been split off from web-api - it is recommended that you Rebuild to this version if you are doing a lot of audio file uploads. Overloading data-api with too many concurrent large file uploads will not longer affect speech recognition sessions in progress.

Backwards incompatibility:

We have changed the result.incrementalTranscript field in the response from GET https://api.voicegain.ai/v1/asr/transcribe/{sessionId} - the new design will better support the changing hypotheses, in particular in commbination with digit and other formatters.

This release fixes the following issues:

#rcj-339: v1/asr/transcribe/{}/transcript?format=text API does not behave like spec w/o interval parameter
#rcj-338: Exception in data-api when upload multiple files concurrently on Edge

Release 1.41.0

This minor release adds the following:

MRCP server’s 3rd-party dependencies get updated to their latest versions
Offline transcription tasks get optimized to increase the throughput
Ability to enable digit formatting as default for a Context
Ability to enable audio capture for all transcriptions in a Context

Release 1.40.2

This maintenance release:

fixes a small bug in EZInit script
improves diarization algorithm
adds alerts in the recognizer

Release 1.40.1

This maintenance release:

fixes issue #rcj-319: diarization setting of minSpeakers:1 and maxSpeakers:2 results in only 1 speaker identified
adds better error handling of invalid request parameters

Release 1.40.0

This minor release adds the following features:

New format parameter to control payload of /asr API callbacks
Test button for quick check of the functionality of an Edge Deployment

Release 1.39.0

This minor release adds the following features:

SIP INVITE for Telephony Bot API supports now also UDP protocol in addition to TCP
Announcements API now supports html format in addition to Markdown (md)
Added ability to download invoices from the Web Console

Release 1.38.0

This minor release adds the following features:

authConf setting for callback authentication
support for new mechanism to specify configuration of services deployed on Edge
invoice API added
changes to user-group API to support attaching groups to Contexts

The release also provides improvements in ML algorithms used for Speech Analytics.

Finally, the release fixes the following issues:

#rcj-315: Missed packet timestamps from dual RTP stream not handled correctly
#rcj-314: user.role is set to User by mistake

Release 1.37.1

This maintenance release:

improves the internal thread Executor setup to make it more resilient
improves on the alerting service to better notify Voicegain support of production issues

Release 1.37.0

This minor release improves the Speech Analytics App.

We added a new type of construct that can be detected in a call. It is called Criteria and allows to conditionally combine NERs, Keywords, Phrases, etc.
Improvements to Call Review view

Our STT API now supports Pause and Mute for real-time transcription.

An improved offline Acoustic Model for English language is also part of this release.

This release addresses the following issues:

#rcj-308: A mic transcription is terminated after 10 minutes
#rcj-305: improve handling for invalid values of fetchTimeout

Release 1.36.3

This maintenance release adds various improvements to the Speech Analytics App (beta).

The release also adds the following enhancements to the API:

max fetchTimeout for getting audio from a URL has been increased from 60 to 90 seconds
it is now possible to set a diarization speaker range starting at 1
GET /sa?detailed=false now returns numIncidents, numChannels, and numSpeakers

Release 1.36.2

This maintenance release fixes one issue and adds one enhancement

#rcj-303: ‘missing websocket name’ error in audiocodes API
#rcj-304: flush RT SA data faster if hypotheses allow it

Release 1.36.1

This maintenance release switches the new account email to AWS SES and it fixes 1 issue:

#rcj-293: differential word cloud empty on Voicegain

Release 1.36.0

This minor release improves the Speech Analytics App.

This release addresses the following issues:

#rcj-288: currentSentiment is all 0 for 30D period
#rcj-286: Call resolution seems to be not calculated correctly
#rcj-285: average Agent Score is incorrect
#rcj-284: latest demo integration gives us empty Differential Word Cloud
#rcj-283: changes to sentiment in the response of GET /sa/call-stats
#rcj-282: changes to how we show sentiment for dashboard
#rcj-281: ensure that output from GET /sa/call-stats uses the subperiod definitions
#saa-13: something wrong with how the sentiment is displayed on the dash
#saa-11: show the value of the sentiment next to smiley on call details page
#saa-10: something wrong with the incident count
#saa-9: tweaks to QA Form read-only view
#saa-8: show Phrases on the analytics page
#saa-7: show tooltip with the Phrase Tag when hovering over highlighted phrase
#saa-6: show PROFANITY keyword in red font
#saa-5: add values of sentiment to the Calls table
#saa-3: Highlight Demo Information part about accuracy
#saa-2: Edited Call Review questions do not persists
#saa-1: wasted space on SA dashboard - 3 bottom cards

Release 1.35.0

This minor release has the following changes:

POST /sa now supports notifying subscribers to a STOMP topic
Added lightweight GET /sa/{saSesId}/status polling method
Added detailed query parameter to GET /sa query method

Release 1.34.0

This minor release has the following changes:

We are releasing a beta of an API compatible with AudioCodes Voice AI Gateway. Once it is in prod we will do integration testing and once AudioCodes confirms compatibility we will make an official announcement about availability.
A completeTimeout setting has been added to the Context settings. This is needed for AudioCodes.
We have added a file upload functionality to the Speech Analytics App. Now you can upload and process your own files in addition to exploring the demo content. You can find the file upload under Settings -> Integrations
We have added a digit formatter to our transcription engine.
The NER accuracy has been improved.
A change to the CC-App which improves usability on mobile phones.

This release also fixes the following issues:

#vgp-793: Ensure that SA config does not allow use of the same tag for a Keyword and a KeywordGroup
#rcj-236: gap value present in the json transcript export
#rcj-261: in phrase detection - location.channel setting not working ok

Release 1.33.0

This minor release has the following changes:

We are switching to AWS SES as our provider for outbound emails. This should makes Password Reset and other emails more reliable.
The Named Entity Recognition (NER) model has been improved.
The beta Speech Analytics App adds: Account time and profile settings, users table and add new users to accounts, full time-zones list.

This release also fixes the following issues:

#rcj-265: validate each email address for POST /user
#rcj-260: in phrase detection - location.time setting should be in seconds
#rcj-259: PII Redaction enhancement (defaults)

Release 1.32.3

This maintenance release has fixes for:

bug #rcj-262: Word Cloud has no content on Demo Integration Dash

Our mail provider has an issue that forces us to temporarily switch to a very simple forgot-password email.

Release 1.32.2

This maintenance release adds two small improvements:

A demo token has been added to control access to Demo Integration in Speech Analytics.
API documentation for Speech Analytics section has been improved.

Release 1.32.1

This maintenance release updates the SSL certificates and also fixes the following issues:

issue #vgp-798: show the incidents on the small audio bar same way as they are shown on the detailed view
bug #rcj-258: Sometimes incidents count not showing on call analysis page
bug #rcj-253: PII Redacted entities shown in audio player view

Release 1.32.0

Limited access Beta release of the Speech Analytics App. To learn more email us at info@voicegain.ai

Release 1.31.1

This release fixes the following issues:

bug #rcj-238: concurrent offline transcription fails due to VAD issue
bug #rcj-237: submitting very short files for diarization fails with ERROR

It also adds a new offline acoustic model with higher accuracy.

Release 1.31.0

New features in this minor release:

Real-time Speech Analytics API - results are provided real-time via websocket
Improved accuracy of the main English language real-time acoustic model
Second English language real-time acoustic model: optimized for use in IVR - provides improved accuracy on long sequences of digits
asr.speechContext parameter to provide hint to the recognizer if a lot of digits are expected
sessions.vadMode parameter to control music rejection - meant to be used mainly for call-center audio to reject music-on-hold
adjusted behavior of the asr.sensitivity parameter - 0 value now corresponds to -40dbFS

This release fixes the following issues:

bug #vgp-775: handle transcript results with no spk assigned to words
bug #rcj-236: gap value present in the json transcript export
bug #rcj-234: Web Console hangs soon after one opens a mic transcription

Release 1.30.1

This maintenance release fixes the following issues:

bug #rcj-228: Noise detected as speech
bug #rcj-227: Real-time transcription of stereo audio files from Amazon S3 fails

Release 1.30.0

New features in this release:

Improved accuracy of the acoustic models - offline and real-time
Improved accuracy of the recognizer in Voicebot scenarios
Ability to set default sensitivity settings from Console Web UI
Returning ERROR instead of NOINPUT if no bytes were received by the recognizer. This makes troubleshooting of RTP streaming easier.

Release 1.29.0

New features in this release:

Support for 2-channel audio streaming via RTP. This is intended mainly for telephony applications where 1 channel is e.g. caller and the 2nd channel is e.g. agent.
An example script that demonstrates this new functionality is on our github.
Added support for MRCP Recognition-Timeout parameter.

Release 1.28.0

New features in this release:

real-time diarization - in addition to off-line diarization which has been supported since October 2020, Voicegain transcribe now support real-time diarization
German language is supported for offline and semi-real-time transcription and recognition
Built-in creditcard grammar supports Luhn check for higher accuracy. Also added Diners Card support.

This release fixes the following issues:

vgp-773 - Incorrect NOMATCH audio zone shown in Transcribe+ for file 4490_pcm_stereo.wav
vgp-773 - Call Metrics broken in Transcribe+
vgp-772 - Female label attached to incorrect speaker
vgp-771 - Left/RIght labels on Transcribe+ (virtual stereo) incorrect
rcj-225 - Review Answers with autopopulated answer choices are not set to the answerValue of the question
vgp-770 - Update NER list
rcj-224 - Comparison method violates its general contract!
rcj-223 - NullPointerException related to a /asr/transcribe request
rcj-222 - Diarization and speaker result in virtual dual channel session displayed on UI is wrong

Release 1.27.1

This maintenance release fixes the following issues:

bug #vgp-769: show informative error message when submitting audio in unsupported/bad format
bug #rcj-219: deleted user account difficult to recover
bug #rcj-218: gaps in audio being played in Telephony Bot API
bug #rcj-216: Call transcripts from Call Sessions not being shown sometimes

Release 1.27.0

New features in this release:

The signup process has been switched from using 2 emails (one from billing and one for the password set link) to using just a single email - it will contain both the info about the billing account created and the link to set the password. This will make the process simpler and reduce the number of password set emails going to a Spam folder.

Release 1.26.0

New features in this release:

Support for “stereo” audio in TWIML audio streaming protocol. This allows for real-time transcription of calls made to Twilio Platform - inbound and outbound channels are transcribed individually. Audio capture (recording) for this audio format is also supported.

Release 1.25.2

This maintenance release fixes the following issue:

bug #rcj-212: Creating new account results in NPE. This resolves the sign-up issue reported earlier.

Release 1.25.1

This maintenance release has fixes for the following issues:

bug #rcj-208: Long audio prompts not played completely in AIVR (Telephone Bot API)
bug #rcj-205: TTS preview not working in Web Console

Release 1.25.0

New features in this release:

New Signup - we have simplified the form
Improved WordCloud - we are using smarter algorithm to generate words and two-word phrases for it. It also now draws faster.

Release 1.24.0

New features in this release:

Spanish language support for off-line transcription. This is beta. Please email us at support@voicegain.ai if you would like Spanish recognizer to be enabled on your account.
PII Redaction has been added to Speech Analytics API. Can be used to remove any chosen Named Entities from the audio and the transcript.
Added support for 3 more new Named Entities that are recognized: ADDRESS, CC (credit card), and SSN.

This release has fixes for the following issues:

bug #vgp-766: Links to Transcript from Context Dash do not work
bug #rcj-196: Microphone transcribe from Web Console is not working on Edge

Release 1.23.1

This maintenance release has fixes for the following issues:

bug #rcj-195: Some short speech audio not being recognized in OFFLINE mode
bug #rcj-194: RTP streaming in 8-bit format fails intermittently

In this release also the accuracy of the real-time and offline acoustic models has been improved.

Release 1.23.0

New features in this release:

Support for RTP streaming fully tested and working - includes ulaw and L16 audio formats
Improved support for simultaneous grammar recognition and large-vocabulary transcription - use of custom language models for large-vocabulary transcription now possible

Release 1.22.0

New features in this release:

Recognition API now able to return results using websockets (previously only transcription API was using websockets for results)
Added support for simultaneous grammar recognition and large-vocabulary transcription.
Offline Speech Analytics API released (beta)
Phone calls made to Telephony Bot API now being processed using Speech Analytics in addition to being transcribed
Improvements in the Language Model feature

Fixes:

bug #rcj-182: certain short audio files (<1.5s) not being transcribed
bug #rcj-180: edits arrive out of order over websocket (issue with ExactTimeCorrectingWordQueue)

Release 1.21.0

New features in this release:

Telephone Bot API now supports SIP INVITE
Transcript table caching in the UI to improve responsiveness
Improved how asr.sensitivity is used for start-of-speech detection

Fixes:

bug #rcj-179: Hints generating misrecognitions
bug #rcj-178: AWS Exception when creating new AIVR app

Release 1.20.0

New features in this release:

improved Edge related web UI
improved Transcribe UI (#vgp-759 - multiple enhancements)
improvements to API around websockets (performance and resource use)
admin tool improvements

Bug Fixes and Enhancements

bug #rcj-164: Certain WAV files with ulaw encoding not being transcribed
bug #vgp-757: Show message when refill attempt fails
enhancement #vgp-761: Enable Delete on expired transcription sessions
enhancement #rcj-158: Terminate sessions which have not received any audio within 30 seconds of being created
enhancement #rcj-157: Terminate real-time sessions with an audio gap of more than 30 seconds

Release 1.19.2

This maintenance release fixes the following bug:

bug #vgp-758: Error when trying to transcribe audio from URL via portal

It also adjusts the CPU and memory settings for a couple of services in order to be able to handle larger loads.

Release 1.19.1

This maintenance release fixes the following bug:

bug #rcj147: Async transcribe result cannot be viewed in cc-app

Release 1.19.0

New features in this release:

Keyword and profanity detection in Transcribe+ page (beta)
Ability to choose real-time or off-line acoustic models for single-GPU Edge deployments
Delete function added to Transcribe page
More accurate real-time acoustic model

Bug Fixes

bug #779: Audio receiving websocket closed too late

Release 1.18.2

This maintenance release fixes the following bug:

bug #768: Problems with accounts with business name containing comma.

and ads the following enhancement:

enhancement #770: add builtin:speech/transcribe as another option to specify transcription in MRCP

The release also includes a higher accuracy offline Acoustic Model

Release 1.18.1

This maintenance release provides fixes for the following bugs:

bug #766: Websocket that returns recognition results is incompatible with OkHttp3

The release also improves the look of transcripts on Transcript+ (beta) page

Release 1.18.0

Major Features and Improvements

Features in this release:

Transcribe+ page (beta)
Switching main url from https://portal.voicegain.ai to https://console.voicegain.ai

Breaking Changes

Old portal url https://portal.voicegain.ai is now a redirect to new url https://console.voicegain.ai

Bug Fixes

bug #764: API requests to incorrect URLs sometimes did not return 404
bug #755: The account without billing permission cannot see transcript result

Known Issues

bug #751: If voice is omitted from <Connect><Stream> request then the audio will play garbled

Release 1.17.0

Major Features and Improvements

Added support for large vocabulary transcription over MRCP
Websocket that returns transcription results is now configurable to be either: a plain websocket, or a websocket that uses STOMP protocol.

Breaking Changes

none

Bug Fixes and Other Changes

since Release 1.16.0

bug #741: Programmable IVR does not wait with start timers till prompt finishes playing to human
enhancement #742: remove underscore from variable values inserted into TTS prompts

Release 1.16.0

Major Features and Improvements

Support for transcription on Twilio Media Streams - provides TwiML <Gather> like functionality
Ability to select different audio channel per session
RTC-API supports large vocabulary transcription
RTC-API supports prompt playback from external URLs
RTC-API input action supports non-bargeinable intro prompt and bargeinable main question prompt
added continuous recognition option

Breaking Changes

audio.channel parameter has been deprecated in /asr/recognize and /asr/transcribe APIs

Bug Fixes and Other Changes

since Release 1.15.1

bug #738: Unable to create accounts starting on digit
bug #736: ‘vars’ not returned when using literal tags in grammar
bug #735: Creating a new broadcast websocket makes the old ones temporarily disappear
bug #733: Prompt text field not returned in events for AIVR Sessions
enhancement #729: Add question.audioResponse.questionPrompt to AIVR Callback response
enhancement #728: Allow AIVR App to send csid either in path or in query or in the body
enhancement #727: add support for start timers in /asr/transcribe
enhancement #726: add DTMF support for JJSGF grammars
enhancement #723: add support for large vocabulary transcription to AIVR
enhancement #722: add to AIVR support for prompts where text is an http URL to the audio to be used
enhancement #721: add support for start timers in AIVR processing loop
enhancement #718: JJSGF tweak - do not require rule names on the left side to be <>
enhancement #716: allow timeout to control the stop of transcription

Release 1.15.1

This maintenance release provides fixes for the following bugs:

#711: Failed to play audio in RTC-API
#709: Adding a clip to clipstore fails on production
#707: Occasional IllegalStateException in Twilio Media Stream

It also is a beta release of the speaker diarization feature.

Release 1.15.0

Major Features and Improvements

Support for speech recognition on Twilio Media Streams via TwiML <Connect><Stream> command
- Voicegain Speech-to-Text integrates with Twilio Media Streams blog post
- How to use Voicegain with Twilio Media Streams blog post
Prompt Manager for dynamic concatenation for prerecorded prompts - can be combined with TTS - currently available only from TwiML <Connect><Stream>
Offline transcription supports output of [music] labels
Improved accuracy from ASR through enhancements to the search algorithm
DTMF detection from audio stream

Breaking Changes

none

Bug Fixes and Other Changes

since Release 1.14.

enhancement #693: Limit duration of microphone transcription to 100 minutes - prevents runaway transcription if someone leaves the microphone on.
enhancement #692: Support externally controlled start of timers on ASR
enhancement #691: Add password strength check to Portal password change page
enhancement #683: Signup will auto-approve all requests for RTC-API feature

Release 1.14.0

Major Features and Improvements

Significantly reduced time for recognition and transcription session setup.
JJSGF now supports tags - both semantics/1.0-literals and semantics/1.0 formats.
Improved accuracy for both offline and real-time acoustic models.

Breaking Changes

Size of inline audio submitted to /asr/recognize or /asr/transcribe has been limited to 4MB

Release 1.13.1

This maintenance release provides fixes for the following bugs:

#661: User logs not accessible
#654: Callback URLs do not support {sessionId}
#651: Microphone transcription on Firefox not working - note: there is still a delay before the first words of the transcript appear in preview

Following enhancements are added:

#657: Account lookup time has been reduced
#656: Log all callback failures to Elastic Search so that users are able to see them

This release includes latest acoustic models with significant improvements to accuracy.

This release also adds minor UI improvements on Transcripts page.

Release 1.13.0

This Minor release focuses on improving websocket streaming.

Major Features and Improvements

Websocket streaming now supports multiple formats, including u-Law 8000 Hz
Added support for PCM Floating-Point 32-bit audio
Added control for recording and transcription to RTC Apps

Breaking Changes

None

Known issues

JJSGF grammar does no support tags yet.
issue #651: Microphone transcription on Firefox not working
issue #607: some customers reported problem uploading some mp3 files

Bug Fixes and Other Changes

since Release 1.12.1

bug #631: No callback received from /asr/recognize
bug #630: Wrong Content-type returned from /data/{uuid}/file
bug #629: Irrelevant parameters returned for audio stream WEBSOCKET
bug #617: Final DELETE callback has the event from previous sequence
bug #616: Transfer reported in RTC Callback misses prompt payback event
bug #573: Websocket streaming locks up due to issues inside 3rd-party library
enhancement #626: Pass value of default vice to RTC Callback
enhancement #625: Set the value of audio.channel in transcription result
enhancement #624: Values of callIsBeingRecorded and toBeTranscribed passed to RTC Callback
enhancement #620: Add to AIVRApp two new fields: recordCalls and percentCallsToTranscribe

Release 1.12.1

This maintenance release provides fixes for following bugs plus provides 2 enhancements:

issue #610: Issues with Listen access to recording and transcript of just ended call
issue #609: Phone number deletion failing
enhancement #613: Disable Delete button for phone numbers used in app - delete would fail anyway
enhancement #611: Provide info about time needed for new RTC App to become active

Release 1.12.0

This Minor release fixes some bugs in the RTC Callback API release last week plus provides some usability enhancements

Major Features and Improvements

TTS preview added - you can no listen to the sound of the TTS voices before making a choice
API Spec document open in full frame - improves navigation within the Spec
Improved usability of User Management

Breaking Changes

None

Known issues

JJSGF grammar does no support tags yet.
issue #607: some customers reported problem uploading some mp3 files
issue #582: Microphone transcription on Firefox gets stuck (never completes)

Bug Fixes and Other Changes

since Release 1.11.1

bug #606: Failed user add was silently ignored
bug #601: Two DELETE callbacks in one session
bug #600: error termination value now supported
enhancement #606: Call Sessions table is by default now sorted in descending order
enhancement #604: Base user role is not explicitly present
enhancement #602: Add sanity stop for runaway RTC session

Release 1.11.0

This Minor release introduces the RTC Callback API for telephony applications of speech recognition.

Major Features and Improvements

New RTC API Application mode – provides tools and APIs to build speech-enabled telephone applications
- purchase phone numbers
- attach phone numbers to RTC Apps
- control flow of you RTC App using Web Callback API
Guide, Help, and API Spec information has been improved.

Breaking Changes

Not really a breaking change but names of two Application Modes have changed:
- Dev API -> STT API (to differentiate from new RTC API)
- ASR -> MRCP ASR

Known issues

JJSGF grammar does no support tags yet.
issue #582: Microphone transcription on Firefox gets stuck (never completes)

Bug Fixes and Other Changes

since Release 1.10.1

bug #566: Link to Terms of Service not visible in Portal
enhancement #585: Alphabetical sorting of contexts
enhancement #583: Direct link to support ticket
enhancement #567: Separate Guide Card for each App

Release 1.10.0

Major Features and Improvements

Added basic JJSGF grammar support (no tags yet)
Changes to Transcription Mode UI
Microphone capture has better compatibility with different browsers
Only a single session is billed if running multiple sessions on same audio input

Breaking Changes

none

Known issues

JJSGF grammar does no support tags yet.

Release 1.9.0

Major Features and Improvements

New API to generate short-lived JWT authentication tokens.
Ability to set allowedOrigins for web API requests to allow cross-origin requests (CORS)
More responsive TTS - part of IVR functionality coming in 2.0.0
Real-time Speech-to-Text price is now 1.25 cent/minute

Breaking Changes

none

Release 1.8.1

This maintenance release provides fixes for following bugs:

#534: Transcript Review page stops spinner before the audio file has finished loading
#530: Last part of incremental word-tree content missing in web api response
#528: GREG chart tool-tip showing incorrect values

Release 1.8.0

Major Features and Improvements

New word-tree format for delivering incremental transcription results.
GREG test grammars can be uploaded in addition to being retrieved from a URL.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

Preemptible flag is now reserved/internal use.

Bug Fixes and Other Changes

since Release 1.7.1

bug #518: Details of tiered pricing not being shown
bug #515: Under-billing for long offline transcriptions
bug #513: In transcript review - speech audio is colored as if it was mere sound
enhancement #524: Allow to set the number of rows on one page for GREG experiment view
enhancement #523: Better GREG Data Upload instruction video

Release 1.7.1

This maintenance release provides fixes for following bugs:

bug #506: Frequency of timestamps in TXT transcript export too high.
bug #505: minimimumDelay setting for Live Transcription websocket has no effect.

With this release the free $1.25 credit for new accounts has been reduced to $0.25 but all accounts now get 600 free speech-to-text minutes monthly.

Release 1.7.0

Major Features and Improvements

Signup has been opened to gmail and other non-business email addresses
Improved acoustic model for offline transcription.
Improved efficiency of the core ASR recognizer.
Many fixes and enhancements around Live Transcription.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

since Release 1.6.3

bug #499: (and related) Session gets stuck in load testing.
bug #492: Unable to enter person’s names with dashes.
bug #488: OutOfMemory error when processing many large requests with inline audio
bug #487: Stopped Live Transcription stuck in Online-Starting state
bug #485: Running Live Transcription with audio.capture flag does not result in working Transcript Preview
bug #481: Secondary output websocket shows controls enabled even though it is not a control websocket
bug #479: Unable to change delay for websocket
bug #474: Unable to change context for websocket
bug #473: “missing websocket name” error when running REAL-TIME together with SEMI-REAL-TIME sessions
bug #469: 504 error when submitting file for transcription - timeout
enhancement #503: IVR-Proxy will now work behind HTTP Proxy.
enhancement #483: Add a label to each Live Transcription session
enhancement #482: Improve archives of Live Transcription
enhancement #478: Improve response of adding a new Websocket

Release 1.6.3

This maintenance release provides an improved acoustic model for off-line transcription.

Release 1.6.2

This maintenance release fixes bug #467

Bug Fixes and Other Changes

bug# 467: Offline transcription of long files sometimes does not report completion - remaining issues have been fixed

Release 1.6.1

This maintenance release fixes the two Issues reported after Release 1.6.0

Bug Fixes and Other Changes

bug# 468: IVR-Proxy not starting.
bug# 467: Offline transcription of long files sometimes does not report completion.

Release 1.6.0

Minor release that adds a much faster offline transcription and a more accurate acoustic model.

Major Features and Improvements

Offline transcription in Edge deployments has been sped up by approximately a factor of 10x.
New acoustic model for offline transcription has significantly improved accuracy.
Previous IVR product has been renamed ASR to better reflect its core functionality. A new product supporting complete end-to-end functionality will be release in June/July.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

since Release 1.5.1

bug #459: Fixed issue that prevented Edge transcription when connection to Cloud was not possible.
bug #450: Fixed issue with resetting expired password.
bug #443: Fixed resource use issue for user log collection.
bug #442: Fixed logout issue from Edge portal.
bug #439: Resolved occasional failure of ASR resource reservation on Edge.
enhancement #460: Cache current balance on Edge to reduce number requests.

Release 1.5.1

Maintenance release that fixes several bugs.

Major Features and Improvements

Added in version 1.5.0:

Web portal is now organized into Voicegain Apps (Transcribe, IVR, etc.) making it easier to navigate the many options by hiding those not applicable to the currently selected App.
Changed New Account Signup form to a wizard style with better explanation of choices available.
Edge deployment now has its own influxDb and Grafana instance.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.5.0 release are listed.

bug #441: Links from Guide Card to support articles fail to login into Zendesk
bug #435: Training mode not supported correctly on Edge Portal
bug #433: New User Wizard (joyride) not reflecting changes in the UI
bug #432: Language Model selection now showing LMs published from other Contexts
bug #431: Billing not tracking all the usage
enhancement #437: Added “Training” option to signup wizard
enhancement #430: Improved icons in Transcribe View playback controls
enhancement #428: Grafana login - auto click OAuth button

Release 1.5.0

Minor release that modifies the Web Portal UI and updates the versions of some core 3rd party libraries.

Major Features and Improvements

Added in version 1.5.0:

Web portal is now organized into Voicegain Apps (Transcribe, IVR, etc.) making it easier to navigate the many options by hiding those not applicable to the currently selected App.
Changed New Account Signup form to a wizard style with better explanation of choices available.
Edge deployment now has its own influxDb and Grafana instance.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.4.1 release are listed.

bug #413: Logs query not respecting date range selection.
bug #393: Insufficient resources allocated to ElasticSearch node.
bug #388: Audio Daemon JVM stats sending to influxdb fails on prod.
enhancement #392: CC-App viewer embedded into Websocket control page.

Release 1.4.1

Maintenance release that fixes several issues and adds minor enhancements.

Major Features and Improvements

Added in version 1.4.0:

Core Functionality
- Web portal now has a built-in log viewer.
- Configurable inactivity timeout has been added.
Web UI
- Many small UI changes improving usability
IVR-Proxy
- Encryption option has been added. Efficiency has been improved. It also now reports version, which makes it easy to spot when you are running an older version and need to update.
Edge Deployment has now been tested and is out of Alpha status.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.4.0 release are listed.

bug #385: Log query error messages not shown in the Web UI
bug #380: Filtering by service name in log query not working
bug #379: MRCP chart not showing data for IVR-Proxy sessions
bug #374: Edge portal has link to MRCP Proxy page
bug #373: Some Edge portal Cloud links point to dev
enhancement #378: Refresh button for MRCP proxy page.

Release 1.4.0

Minor release adding features and improvements as well as bug fixes.

Major Features and Improvements

Added in this version:

Core Functionality
- Web portal now has a built-in log viewer.
- Configurable inactivity timeout has been added.
Web UI
- Many small UI changes improving usability
IVR-Proxy
- Encryption option has been added. Efficiency has been improved. It also now reports version, which makes it easy to spot when you are running an older version and need to update.
Edge Deployment has now been tested and is out of Alpha status.

Known issues

Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.3.3_beta release are listed.

bug #357: https://portal.voicegain.ai/billing url fixed
bug #338: Unable to control certain broadcast websockets
bug #337: Websocket detail page blank for certain websockets
bug #336: Audio Sender link downloads incorrect configuration if name mistyped
enhancement #367: SSO sends directly to destination if only one destination available
enhancement #362: Format of “My Status” card has been improved
enhancement #349: MRCP chart shows data specific to Context in which it is viewed
enhancement #342: Broadcast Websockets can now be deleted
enhancement #335: Add option to choose encryption for IVR Proxy at download link

Release 1.3.3_beta

Maintenance release that fixes several issues and adds minor enhancements.

Major Features and Improvements

(Info carried over from 1.3.0_beta)

Core Functionality
- Support for multiple Acoustic models - may be configured in ASR Settings for Contexts and for each API request
IVR
- Added to GREG Tool ability to run Experiments on external ASR. (1.3.1_beta adds preview of new GREG interface)
- Added downloadable IVR-Proxy for interfacing local MRCP Clients to Voicegain ASR in the Cloud.
Transcription
- Redesigned the Transcript Review UI.
- Added highlighting of non-speech sound and unrecognized speech.
- Added microphone input capture.
Other
- Added Voicegain sign-up dialog form (previously new accounts had to be created using Admin tool).
Speech.Works
- Transcript Review page has been changed to be the same as the new page in Voicegain Portal

Known issues

Edge deployment is still in Alpha status. It is functional for Trial use. Can be enabled upon Customer request.
Audio Sender Daemon does not support encryption yet.
Grafana Dashboards configurations are very basic
issue #322: OAuth sign-in to Grafana does not work if a session has been open for a while

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.3.2_beta release are listed.

bug #321: Error when creating a new account with business name containing a hyphen
bug #317: Deleting user via Admin Console does not work
bugs #314,316: Grafana - uses can see the main dashboard, but cannot see their Account dashboards
enhancement #320,325: Modifications for Sign-Up form and page (incl. email validation)
enhancement #318: Add IVR chart for MRCP messages

Release 1.3.2_beta

Maintenance release that fixes several issues and adds minor enhancements.

Major Features and Improvements

(Info carried over from 1.3.0_beta)

Core Functionality
- Support for multiple Acoustic models - may be configured in ASR Settings for Contexts and for each API request
IVR
- Added to GREG Tool ability to run Experiments on external ASR. (1.3.1_beta adds preview of new GREG interface)
- Added downloadable IVR-Proxy for interfacing local MRCP Clients to Voicegain ASR in the Cloud.
Transcription
- Redesigned the Transcript Review UI.
- Added highlighting of non-speech sound and unrecognized speech.
- Added microphone input capture.
Other
- Added Voicegain sign-up dialog form (previously new accounts had to be created using Admin tool).
Speech.Works
- Transcript Review page has been changed to be the same as the new page in Voicegain Portal

Known issues

Edge deployment is still in Alpha status. It is functional for Trial use. Can be enabled upon Customer request.
Audio Sender Daemon does not support encryption yet.
Grafana Dashboards are not configured yet

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.3.1_beta release are listed.

bug #312: Wrong URL returned for Audio Sender bootstrap download link
bug #311: Expiry time for Audio Sender bootstrap download link is shown in UTC
bug #309: Wrong websocket url returned for real-time transcription on Prod
bug #306: Stop words shown in word cloud for Transcript Review page
bug #305: After sign-up from the form account web name is derived from owner name instead of company name
bug #304: No ASR settings visible if account has no premium features
enhancement #320: Sign-up form has been modified
enhancement #313: Redeploy button has been added to Edge Deployment page
enhancement #307: Refresh button added to ASR Availability card
enhancement #301: Improved Wizard for EZSetup of Edge Deployment
enhancement #298: In new GREG show audio hash instead of id and allow copy of either
enhancement #295: Better file names for downloaded transcripts

Release 1.3.1_beta

Maintenance release that fixes several issues and adds minor enhancements.

Major Features and Improvements

(Info carried over from 1.3.0_beta)

Core Functionality
- Support for multiple Acoustic models - may be configured in ASR Settings for Contexts and for each API request
IVR
- Added to GREG Tool ability to run Experiments on external ASR. (1.3.1_beta adds preview of new GREG interface)
- Added downloadable IVR-Proxy for interfacing local MRCP Clients to Voicegain ASR in the Cloud.
Transcription
- Redesigned the Transcript Review UI.
- Added highlighting of non-speech sound and unrecognized speech.
- Added microphone input capture.
Other
- Added Voicegain sign-up dialog form (previously new accounts had to be created using Admin tool).
Speech.Works
- Transcript Review page has been changed to be the same as the new page in Voicegain Portal

Known issues

Audio-Sender bootstrap download not fully configured yet. Will make an announcement when it is ready.
Edge deployment is still in Alpha status. It is functional for Trial use. Can be enabled upon Customer request.
Audio Sender Daemon has to be deployed manually. It also does not support encryption yet.

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.3.0_beta release are listed.

bug #280: Prevent opening transcription viewer in transcription is in error state
bug #283: SSO to Grafana analytics dashboard does not work
bug #284: MRCP and IVR-Proxy suffer packet loss under concurrent load
bug #286: Web-api in on-prem environment should not attempt to talk to Billing API directly
bug #292: Data usage not monitored propertly due to missing Cloud Function Permission
enchancement #270: Modified text of the Billing emails to add clarity
enchancements #275,277, 278: Add Corpus dialog specifies expected file type, size, storage cost, and prevents upload of files that are too large
acoustic model: VoiceGain-rt-ivr-en-us:13 (reduced error from 0.0857 to 0.0831)

Release 1.3.0_beta

Minor release that adds extra IVR and Transcription functionality.

Major Features and Improvements

Core Functionality
- Support for multiple Acoustic models - may be configured in ASR Settings for Contexts and for each API request
IVR
- Added to GREG Tool ability to run Experiments on external ASR.
- Added downloadable IVR-Proxy for interfacing local MRCP Clients to Voicegain ASR in the Cloud.
Transcription
- Redesigned the Transcript Review UI.
- Added highlighting of non-speech sound and unrecognized speech.
- Added microphone input capture.
Other
- Added Voicegain sign-up dialog form (previously new accounts had to be created using Admin tool).
Speech.Works
- Transcript Review page has been changed to be the same as the new page in Voicegain Portal

Known issues

Audio-Sender bootstrap download not fully configured yet. Will make an announcement when it is ready.
Edge deployment is still in Alpha status. It is functional for Trial use. Can be enabled upon Customer request.
Audio Sender Daemon has to be deployed manually. It also does not support encryption yet.

Breaking Changes

none

Bug Fixes and Other Changes

Only bugs present in the previous 1.2.2_beta release are listed.

bug #190: Incorrect parsing of urls with websocket names containing dot ‘.’
bug #203: Gateway timeout when creating a Language Model
bug #206: Offline transcription does not report certain Errors (stuck on Fetched)

Release 1.2.2_beta

Maintenance release in order to fix 2 breaking bugs.

Major Features and Improvements

None in this maintenance Release.

Known issues

Edge deployment is still in Alpha status. Disabled by default. Can be enabled per Customer Account upon request.
Cloud IVR has not passed full test suite yet. Disabled by default. Can be enabled per Customer Account upon request.
Audio Sender Daemon does not support encryption yet.

Breaking Changes

Broadcast Websocket functionality not longer part of the default set of features.

Bug Fixes and Other Changes

bug #185: Unable to generate JWT from Portal (Context Web Settings not visible in Web UI)
bug #181: Websockets control page not visible in Web UI (even for accounts with the websocket broadcast feature)

Release 1.2.1_beta

Maintenance release in order to fix issues with Preemptible GPU instances.

Major Features and Improvements

Flexible mechanism to allocate various types of GPUs to ASR instances. Circumvents the problem with certain GPUs temporarily not being available in the Cloud.

Known issues

Edge deployment is still in Alpha status. Disabled by default. Can be enabled per Customer Account upon request.
Cloud IVR has not passed full test suite yet. Disabled by default. Can be enabled per Customer Account upon request.
Audio Sender Daemon does not support encryption yet.

Breaking Changes

Broadcast Websocket functionality not longer part of the default set of features.

Bug Fixes and Other Changes

bug #178: NullPointerException caused by an invalid request for /asr/recognize
bug #177: Fixed incorrect Uptime computation for Voicegain ASR
enhancement #176: Added Announcements on home page

Release 1.2.0_beta

This release adds fine-grained control of features available to various types of accounts.

Major Features and Improvements

More convenient Billing status info and control from within the Portal.
Fine-grained control of features enabled for each account. This allows Voicegain to control access to features that still are still in limited relase to certain accounts only.
Enhanced Admin/Account Provisioning Portal - password reset, referrer/referral, premium features, etc.

Known issues

Edge deployment is still in Alpha status. Disabled by default. Can be enabled per Customer Account upon request.
Cloud IVR has not passed full test suite yet. Disabled by default. Can be enabled per Customer Account upon request.
Audio Sender Daemon does not support encryption yet.

Breaking Changes

none

Bug Fixes and Other Changes

bug #167: Release Notes not being displayed.
bug #170: Fixed slow redraw of modified words in CC-App
bug #171: Login failing due to case difference is user email
enhancement #173: Favicons different between various apps and prod/QA/dev
enhancement #174: CC-App supports Account landing page

Release 1.1.0_beta.2

This release adds functionality that did not make it into the initial release. It also fixes several bugs.

Major Features and Improvements

Added ASR engine resource status monitoring to Voicegain Portal UI
Added Password strength checking
Added Release Notes info to Voicegain Portal UI
All three billing styles (manual-refill, auto-refill, invoice) now supported for Enterprise Voicegain customers. Speech.Works customers are created with manual-refill style.
Provisioning&Admin Portal shows 3 new account fields: referrer, referral, billingStyle.

Known issues

Edge deployment is still in Alpha status - please contact us before attempting Edge Deployments
Cloud IVR has not passed full test suite yet. Also, it lacks documentation. Please contact us if you would like to test Cloud IVR.
Audio Sender Daemon does not support encryption yet.

Breaking Changes

none

Bug Fixes and Other Changes

Most of the bugs fixed in this release are related to login/authentication and to working with Language Models.

bug #142: Provisioning Portal times out after login.
bug #143: ASR resets occasionally in the middle of transcription. Was due to empty request to NLM.
bug #144: Password lockout was mistakenly set to occur only after one mistake. Has been increased to 5.
bug #145: 401 errors when logging in after a period of inactivity
bug #147: Service name at the end of the URL in the password (re)set email incorrect.
bug #148: New Language Models not shown unless user has Manager role
bug #149: Language Model created as Built-in shows up as User model
bug #153: Bad redirect from Password Set Dialog
bug #158: Language Model card for a Context is not showing all the models
bug #160: Building a Built-In model fails
bug #161: Ensure correct results are returned from Language Model Query (built-in models, etc.)
bug #162: Transcription with built-in Language Models fails at initial checks
bug #163: Offline transcription with built-in Language Model not working
minor bugs #150, #157

Release 1.0.0_beta.1

This is a first public release of Voicegain Speech-to-Text Platform.

Major Features and Improvements

This being a first public release, here is a list of key features of the Voicegain platform:

Cloud based Speech-to-Text API supporting both Transcription (large vocabulary) and Recognition (limited context-free grammars)
MRCP/VXML/GRXML compatible ASR - tested with Dialogic and Aspect (Voxeo) VXML platforms
Real-Time transcription - core Real-Time Speech-to-Text engine plus tools like: Audio Sender, Web UI for managing multiple transcription channels, Web based client for displaying live transcriptions
Voicegain Enterprise Web Portal
Simplified Transcription only Web Portal (Speech.Works)

Known issues

Edge deployment is still in Alpha status - please contact us before attempting Edge Deployments
Cloud IVR has not passed full test suite yet. Also it lacks documentation. Please contact us if you would like to test Cloud IVR.
Audio Sender Daemon does not support encryption yet.

Breaking Changes

none

Bug Fixes and Other Changes

We will begin reporting bug fixes starting from the next release