Heya -- I built this mostly using openhands... just been curating a ton of research for the past 1.5 years. I've been putting together a RAG Production Guide based on real enterprise deployment experience — covering chunking, retrieval tuning, evaluation, observability, and governance considerations that actually matter when you're running RAG in prod (not just in a notebook). It's open source and I'd love this community's eyes on it. You all will catch things I missed. 🙏 What I'm asking for: ⭐ Star it if you find it useful 🍴 Fork it and build on it 💬 Drop feedback here or open an issue/PR directly 👉 https://github.com/MarioLazo/rag-production-guide Biggest gaps I know I have: more on evaluation frameworks and pipeline monitoring. If that's your wheelhouse, would love a contribution or even just a convo. What are you seeing in your RAG deployments that should be in here?

Wow! you have a lot there. I have worked on RAG for a while and here some some quick thoughts to consider adding: • Extraction out of documents - page layout, captioning • Importance of Metadata - more information makes retrieval easier • Rise of Agentic Search - its not one shot any more I have a long blog post and links to some videos on RAG i have done over here: https://rajivshah.com/blog/rag-agentic-world.html

good points!!!! this is very helpful