https://docs.llamaindex.ai/en/stable/examples/retrievers/bm25_retriever.html https://docs.llamaindex.ai/en/stable/examples/retrievers/bm2... Also maybe try to

https://docs.llamaindex.ai/en/stable/examples/retrievers/bm25_retriever.html https://docs.llamaindex.ai/en/stable/examples/retrievers/bm2... Also maybe try to include tags or categories when you index and then you can filter on those when doing the vector search. Might get a similar effect from BM25. Also llamaindex does RAG better than some other solutions.