Updated on May 15, 2025
In April, we released our integration with Twilio. We also focused on improving our voice AI experience to improve your customer service workflows.
Features
Twilio Integration
You can now use Twilio to handle your incoming support requests with an AI agent. You just need to add the webhook URLs to your preferred support number on Twilio. For more instructions on how to integrate, watch our video or read our blog.
Improvements
- We have added the ‘Extract Entities’ functionality to our Anthropic and Gemini-based chatbots.
- We have configured events for voice AI to improve call and performance tracking.
- Now, when you train a URL that has a link to a PDF, DOCX or CSV file, your AI agent will automatically read and train itself on the file.
Bug Fixes
- We fixed a TypeError where undefined is not iterable — this error occurred when the system attempted to read a value using Symbol.iterator on an undefined object.
- We updated how First Response Time (FRT) is calculated. It will now only begin after a conversation is assigned to an agent, ensuring more accurate reporting.
- We resolved an API error that occurred when trying to delete bots, ensuring smoother bot management.
- We fixed a UI issue where the interface would break if users responded via email instead of the chat widget.
- We corrected a bug where conversations were being assigned to agents who had been deleted, leading to misrouted tickets.
- We fixed an issue where the bot was responding with “No records found” instead of using the default fallback message, improving the user experience during unmatched queries.
