You’re offline. This is a read only version of the page.
9
I am curious to know how we can clean the response before sending it when using conversational boosting. I would like to strip out any references to citations before sending a response to the user. However, this is surprisingly daunting as it seems like isMatch, Match, and MatchAll are not permitted in CoPilot Studio.
It seems like perhaps there may be a way to do this using entities, as they have built-in regular expression support, however, I can't seem to find a way to feed them the AI response, rather, than prompting the user for input.
Any thoughts or help with this would be greatly appreciated.
For additional context, we do not want to expose users to our documents as they are being cut up to meet the 3MB quota, in addition, some of our documents context would be irrelevant or confusing as video transcripts, code snippets, and archived announcements would require additional context, formatting, and multimedia support in order to be relevant.
Share
Report
All responses (
Answers (
