2024-09-01 Changelog
This update brings new features, bug fixes, and improvements to enhance the performance and usability of the platform.
New Features:
- JWT Token Restrictions: You can now restrict the JWT token used for authentication in the stand-alone version of the Transcript Editor to a specific transcript using custom permission claims. This adds an extra layer of security.
- Mandatory Exp Claim for JWT: The "exp" (expiration) claim in the JWT token is now mandatory for the stand-alone version of the Transcript Editor. This ensures tokens expire properly, improving security.
Bug Fixes:
- Paragraph Order in Transcript Editor: Fixed an issue where paragraphs in the Transcript Editor were not saved in the correct order.
- Voice Activity Detection (VAD) Empty Segments: Fixed a bug where empty segments from VAD were causing the entire Speech Recognition job to fail.
- External URL Downloads: Improved the reliability of downloading external URLs as job sources by adding retry logic and optimizing how timeouts are handled.
Improvements:
- Dictionary Specification Threshold: The default threshold for Dictionary Specification has been increased to 0.90. This improves the accuracy of word matching.
- Fuzzy Dictionaries for Speech Recognition: Enhanced fuzzy dictionary functionality for better word correction during Speech Recognition.
- Minor Bug Fixes: Additional small bug fixes to improve platform stability and performance.
All Updates:
Click here for all detailed informations and changes in our Technical Documentation
Changelog Newsletter:
We have a dedicated Newsletter with only Technical Changes and Updates, click here to subscribe via LinkedIn or via E-Mail.