SUVA v23.10 Release Notes

SUVA's October release (v23.10) is available for all users starting on October 31st. This new release marks another milestone toward the continuous improvement of SUVA, ensuring a better experience for admins and end-users alike.

SUVA v23.10 introduces a new feedback mechanism for SUVA admins to gather user feedback on chatbot-generated responses. Additionally, SUVA-powered chatbots have been enhanced to return clearer, more concise, and more readable responses.

The details of the key highlights of SUVA v23.10 release are as follows:

Feedback Per Response

Feedback per response is an approach where the chatbot solicits feedback from the user after answering each client query. This mechanism is primarily designed for LLM-powered responses to analyze their effectiveness.

This mechanism improves the chatbot interactions quality and performance by continuously gathering information about the user's satisfaction, understanding, or preferences.

For a traditional chatbot, when LLM integration is not active, the response mechanism remains unchanged. A response is only sought at the end of a node or story.

Related Doc. An Overview of the SUVA Chat Window

Feedback Report in Analytics

A new Feedback report has been added to SUVA Analytics where admins can review user feedback on chatbot responses. The feedback report is vital for enhancing the chatbot's performance and user experience.

Related Doc. Feedback: Review User Reactions on Chatbot Conversations

Responses with Lists in LLM-Powered Chatbot

SUVA's LLM-powered chatbot now has the capability to analyze user context and the nature of the information being conveyed, and present the information in sequential manner for better readability. Using bullet points and numbered lists in chatbot responses isn't just about formatting; it's about improving the overall user experience.

Note. The list responses are generated for 'how-to', 'steps, 'comparison', 'highlights', and similar kinds of queries.

Bug Fix

A hyperlinking issue has been resolved in this release. In some chatbot responses, the URLs showed instead of hyperlinked texts. Users saw https://suva.ai/ instead of SUVA Official Site. Now, this issue has been fixed.