Feature Update: Conversation AI Bot Response Feedback and Fine-Tuning

With this enhancement, we aim to empower our users to have more control over their bot’s responses and improve the overall conversational experience.

We have implemented a comprehensive Feedback loop that works seamlessly with the Auto-reply mode (in private beta) and the Bot trial chatbox.

# Bot Trial tab in Conversation AI Settings:

– On the Bot Trial Page, you will now discover “thumbs up” and “thumbs down” icons next to each bot response.

Thumbs Up: By clicking the “thumbs up” icon, users can provide positive feedback, indicating that the bot’s response is correct.

Thumbs Down: Selecting the “thumbs down” icon opens a pop-up window where you can edit both the question and the answer provided by the bot. This empowers users to make corrections or improvements to either the question or answer.

– After making necessary edits, users can click “Train Bot” to add the updated Q&A pair to the bot’s FAQ database (customise bot responses). This action will not only improve the bot’s responses but also enhance its future interactions.

# Conversation Page (for Auto-Pilot’s private-beta customers):

– The Conversation Page now mirrors the functionality of the Bot Trial Page.

– Users can utilize the “thumbs up” and “thumbs down” icons to provide feedback or make corrections as needed.

# Contact Details Page (for Auto-Pilot’s private-beta customers):

– The Contact Details Page also follows the same feedback mechanism as the Bot Trial Page.

– Users can use the “thumbs up” and “thumbs down” icons to provide feedback or make corrections as needed, ensuring a consistent feedback loop and correction process across different sections of our platform.

# Benefits:

Improved Bot Responses: Users can now actively participate in refining their bot’s responses, leading to more accurate and contextually relevant interactions.

Effortless Training: The “Train Bot” feature simplifies the process of updating the bot’s knowledge base, allowing it to learn from user feedback.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top