October 6, 2025

​

Improvements

• [Enterprise] Enhanced voice pronunciation with automatic currency conversion for more natural speech patterns. Read more about it here!
• Audio-based variable extraction is now available to improve accuracy. The model processes audio directly instead of relying on an intermediate speech-to-text step, using richer signal to extract variables. Enable this under Advanced Variable Extraction Settings in your pathway nodes.

• Fixed call filtering to include calls not linked to any pathway when using “NOT” operator filters
• Enhanced call log export functionality with support for additional filtering options and improved reliability
• Added pre and post transfer duration fields to call export data for better call analytics
• Refreshed call logs interface with improved visual design for better user experience

View call logs interface improvements

• Improved route condition handling with automatic parsing of boolean and null values for more reliable pathway logic
• Improved custom components visibility on web widget start nodes for consistent display
• Resolved Pathway topbar overflow issues for a cleaner, more consistent interface

View pathway topbar improvements

• Resolved voice overlay display issues in pathway chat interface
• Improved error handling when deleting pathways to prevent unexpected failures

• Added maximum call duration configuration in persona settings for better call management

• Clearer visuals for uploaded files in the knowledge base, including better file type identification

View knowledge base file type improvements

• Resolved issues with SMS number configuration updates across Pathway and Objective settings
• Added request data variable support to SMS testing for more comprehensive coverage
• Enhanced SMS filtering with “last message at” timestamp filtering