-
Notifications
You must be signed in to change notification settings - Fork 39
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
AI Assistant: Get feedback on AI responses #2495
Comments
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This needs more design detail. But this is required before general release (especially if we're not going to post chat logs to slack yet) |
The ideal solution here for me is to allow users to flag a response from the assistant. We should have a text under the responses that say, flag this response and on click, it should open a modal that asks the user to select the reasons(s) for flagging the response- incorrect information, incomplete answer, irrelevant answer, confusing or unclear, too long or verbose answer, too short or lacking detail, outdated information, not based on documentation, formatting issues, and other. This "flag this answer" feature should only show when a user hovers over the answer from the assistant. When flagging a response, users can select more than one reason and should also provide more details about their reason for flagging the response. The responses will be saved in the db (maybe in a new table). |
@josephjclark what do you think about this? |
Here is how Claude does it (same as your proposal @christad92), and I think it's really cool. Click on these For positive feedback For negative feedback |
@christad92 to investigate whether we can attach the AI-chat session ID to a posthog survey. Then we can have a really simple flow like:
|
Here are a few things we want to do:
The text was updated successfully, but these errors were encountered: