More configuration of RAG retrieval pls

The plug-in you have for open AI file storage is a nice start, but it doesn’t expose many of the possible parameters that I want to control. For example, I want to be able to identify and return just the top 20 trunks without interpretation along with document, sourcing, timestamps, etc..

At the moment we’re getting basic generation of what’s retrieved by the rag query, but we can’t control it.

This would be a game changer for me