due to randomness of the model itself, the best way to achieve stable results is to use RAG technique. In your case, instead of relying only on the model to find relevant documents, you could use a Sharepoint action like `Get files` - when invoked by the agent, it could get all files and filter out only the ones that the user asked for. API call will always return all results based on the query.
In case of any other questions, let me know. If the answer is correct, mark it as a solution, so that others can benefit from it.
Best regards,
Artur Stepniak
Was this reply helpful?YesNo
Under review
Thank you for your reply! To ensure a great experience for everyone, your content is awaiting approval by our Community Managers. Please check back later.