Version 5.13

Version 5.13

July 25, 2025

Enhancements

Azure AD Authentication for SSO
Azure Active Directory is now supported for Single Sign-On (SSO) in cloud environments. This enhances security and simplifies user access management by integrating enterprise-level identity providers.

Monthly SR Limit Per Agent
Administrators can now assign a monthly speech recognition (SR) duration cap per agent 150 hr/agent. This control helps optimize resource consumption and cost management across tenants.

Waveform Disabled for Very Long Conversations
Waveform rendering is now automatically disabled for conversations longer than 75 minutes. This improves UI performance and prevents potential rendering issues for long audio sessions.

Audio File Compression Using OPUS
A new feature allows compressing audio files into the OPUS format, significantly reducing storage requirements while preserving audio quality.

Hold Count Parameter in Categories
Categories now support filtering based on the number of "Hold" actions within a conversation. This enables more precise categorization based on customer experience metrics.

Opening New Index for All Tenants
New index structure is enabled on all tenants

Flexible LLM Configurations per Task
Large Language Model (LLM) configurations can now be customized per task (e.g., categorization, Q&A). This allows for more accurate results by using optimal model settings for each GenAI function.

Saving AI Matching Info After Re-Training
AI-generated category matching metadata is now retained even after model re-training using the same topics. This improves traceability and analysis history.

Analysis Completion Date Added to CA_ConversationDetails
Each conversation now records the date and time when its analysis is completed. This improves auditing, filtering, and processing traceability. It can be used in receiving analysis results to another system.

AI Generated Categories with LLM Configurations
AI-generated categories now benefit from tailored LLM configurations, enhancing accuracy and flexibility based on tenant-specific use cases.

Aligning Redaction Logic for AI Categories
Redaction rules for AI-generated categories are now aligned with those for manually created categories, ensuring consistent privacy enforcement.

Pause/Stop Inference Models
Admins now have the ability to pause or stop active inference models in real time. This provides greater control over system performance and cost.

Coaching & Training as a Feature
A new feature set for coaching and training has been introduced. It allows organizations to assign, track, and manage performance improvement sessions within the platform.

Direction Filter in Statistical Pages
Statistical Comparison and Statistical Analysis pages now include a direction filter (inbound/outbound), giving users more precise control over their data views.

Improvements

Updated Terminology: 'Block' to 'Monologue'
The term "Block" has been fully replaced with "Monologue" across the form details and waveform views to better reflect the uninterrupted speech units.

Enhanced Pronunciation Support
The pronunciation editor now supports additional special characters and allows copy-pasting into multiple fields at once. This makes it easier to manage domain-specific pronunciation dictionaries.

Minor UI Enhancements
Several minor improvements have been implemented to enhance the user interface. These enhancements make the interface more intuitive and user-friendly, contributing to a better overall user experience.

Minor Bug Fixes
Various minor bugs have been addressed to improve overall system stability and performance. These fixes result in a smoother and more reliable user experience.