Important Note
Please don't forget to check the version history of LMS as it is a system that allows users to store speech recognition models.
Q4 2025 (Oct–Dec 2025)
Enhancements
- Reduced Storage Costs with Automatic Audio Compression
Audio recordings captured by SR are now automatically compressed before long-term storage, significantly reducing disk usage and network transfer size for logs and test data. On-premise customers who need higher fidelity can raise the compression quality via configuration, and files can still be converted back to their original format when needed for download or analysis.
Bug Fixes
-
Fixed a Database Permission Issue Affecting SR on MS-SQL
Resolved a partition permission issue that could affect SR deployments running on MS-SQL databases. -
Fixed a Processing Delay in Dolphin Context Biasing
Resolved a processing time issue affecting context biasing in the Dolphin recognition engine.