What should a production-ready RAG system add?
The RAG project already includes several advanced search features. It combines BM25, which finds matching words, with , which finds text with similar meaning, and then blends the results with . It uses Cohere reranking to reorder results and HyDE query expansion to make the user’s question easier to match with useful documents.
It keeps a saved , updates only new or changed content, adds , and includes a no policy. The main issue is what else is needed before this can be considered , instead of just a more polished version of a basic RAG demo.
Key points
- The design already uses BM25 and together.
- merges search results, while Cohere reranking changes their order.
- HyDE query expansion tries to improve the question before searching.
- The system keeps a saved index, updates it incrementally, cites sources, and aims to avoid unsupported answers.
- The goal is to move beyond a basic RAG demo and make it ready for real use.