We've rolled out targeted performance and reliability improvements across the platform to make things faster, more stable, and easier to manage.
Database query & load optimizations
Eliminated redundant DB operations across ~70% of system queries, significantly lowering database load and boosting overall efficiency.
Faster endpoints
Reduced execution time for frequently accessed, high-traffic endpoints — responses are noticeably quicker.
Responsiveness & reliability upgrades
Platform hardening to improve responsiveness and stability under heavy load.
Conversation controls & agent visibility (IntraChat)
Added a Delete Conversation action and now show the agent name at the top of each conversation for clearer context and management.
What’s new: remove individual conversations from the chat UI and instantly see which AI agent handled the exchange.
Scheduler & auto-scrape / training fixes
Improved scheduler reliability for automated scraping and training jobs.
Fix highlights: resolved jobs that failed to start or finish, ensured training triggers after scraping, and improved scheduler log output for easier debugging.
Per-message token visibility
Token counts (input/output) are now shown per message in the dashboard widget so you can monitor usage and optimize model consumption.
Conversation ID in custom function payloads
Custom function calls now include the Conversation ID (UUID) in the payload for precise tracking and simpler correlation between API events and chat context.
Leads tab — test & notification history
Improved visibility for lead-capture workflows with a Test Notification option and an Email Notification History to track when alerts were sent and what triggered them — simplifying testing and troubleshooting.