Curious if you'd be good enough to go with a RAG [1] approach where you search relevant text and pass it to the llm when answering a question. I've been working on a simple tool for that called Formualic thay enables RAG workflows! Found at https://formulaic.app