The ability to view and edit AI response information directly from the conversation view enhances the overall transparency, control, and accuracy of AI-generated responses. By offering a quick and easy way to understand and modify the AI's components, this feature ensures that users can refine their AI’s performance in real-time. This article provides an overview of the AI Response Info functionality, its benefits, and a step-by-step guide on how to access and edit it.
Transparency: View the underlying components of AI-generated responses, including prompts, intents, and data sources.
Control: Edit FAQs and website information directly from the conversation view to ensure that responses are accurate and up-to-date.
Efficiency: Make real-time updates to knowledge sources without navigating away from the conversation, improving the AI's performance.
Instant Updates: Changes made to the knowledge base are immediately reflected in AI responses, improving the accuracy of future interactions.
Chat History Tab: Access past conversations that contributed to generating the AI response.
Action Execution History: Monitor all actions triggered by the AI to generate its response, enhancing transparency and troubleshooting.
AI Response Info allows users to view the following components for any AI-generated message:
1. Navigate to the Conversation View:
Open the conversation where AI-generated messages are present.
2. Identify the AI Message:
Look for the AI-generated message in the conversation. An AI icon will appear below the message to indicate that it was generated by the AI. The new window displays all relevant AI response details, including:
Response: The AI's reply to the user. Users can see the Agent Name in the AI Message Details as well. This response can be edited to adjust how the AI responds in future conversations, but it does not change the message that was already sent.
Prompt: The instructions the AI followed to generate the response. This can be edited in the AI Response Info sidebar to modify future responses.
Intent: The goal or purpose behind the response. While this cannot be changed directly in the sidebar, adjustments can be made via the Conversation AI settings.
Sources (Knowledge Chunks & FAQs): The knowledge sources used to generate the AI's response. This could include URLs or entries from the bot's knowledge base, such as FAQs or other data.
Chat History Tab: This lets users view previous conversations that were considered when generating the AI's response. This ensures that only the relevant interactions between the bot and the user are displayed, providing clear context.
Action Execution History: Below the Chat History tab, an expandable section shows all the actions the AI took to generate its response, such as booking an appointment, stopping the bot, updating contact fields, or triggering a workflow. It includes:
Action Description: What the AI did (e.g., retrieving appointment slots).
Request Sent to AI: The specific request made to the AI (e.g., "Getting slots on 14th October from 7 AM to 7 PM").
Response Generated by AI: The AI's final output (e.g., "Slots fetched from the calendar").
3. Save Changes:
Any changes made will be saved and automatically reflected in the AI’s future responses.
FAQ:
What is AI Response Info?
AI Response Info allows you to view and edit the components of AI-generated responses, such as the response itself, prompts, intents, and data sources. This feature provides transparency and control over how the AI interacts with users.
How can I view the AI Response Info in a conversation?
To view the AI response information, open the conversation where the AI-generated message is present. Click on the AI icon to access the details of the response, including the prompt, intent, and sources.
Can I edit the AI-generated response?
Yes, you can edit the AI's response to modify how it will reply in future conversations. However, the changes won’t affect messages that have already been sent to the user.
What is the purpose of the Chat History Tab?
The Chat History Tab allows you to view past conversations that were considered by the AI when generating the current response, ensuring relevant context is provided.
What information is available in the Action Execution History?
The Action Execution History shows all the actions the AI took to generate the response, such as booking an appointment or triggering a workflow. It includes the action description, the request sent to the AI, and the response generated by the AI.
