Hi, and thanks for maintaining ai-engineering-hub.
The collection of tutorials on LLMs, RAG, and agents is very helpful.
I wanted to suggest one possible addition for the RAG side, focused on debugging and failure analysis.
Project
What it is
- A 16 mode failure map for RAG and LLM pipelines, with modes numbered No.1 to No.16.
- Each mode has a markdown page that describes:
- how the failure looks from the user side,
- what usually went wrong in the system,
- and a minimal structural fix.
- There is a small text based debugger where engineers paste a bug description and receive a Problem Map label plus the best doc to read first.
Why it might fit this repo
- Many notebooks here show how to build working RAG systems. WFGY tries to answer “what went wrong when it quietly fails in production”.
- It could be linked in any RAG debugging or “common pitfalls” section as a concrete map that students can follow.
- Since it is all markdown and prompt based, it does not lock people into any specific tool.
If this is in scope, I would be happy to help draft a short paragraph or a small notebook that shows how to use the Problem Map when debugging RAG experiments.
If you prefer to keep the hub focused on your own content only, that is completely fine.
Thanks again for sharing these materials.
Hi, and thanks for maintaining ai-engineering-hub.
The collection of tutorials on LLMs, RAG, and agents is very helpful.
I wanted to suggest one possible addition for the RAG side, focused on debugging and failure analysis.
Project
What it is
Why it might fit this repo
If this is in scope, I would be happy to help draft a short paragraph or a small notebook that shows how to use the Problem Map when debugging RAG experiments.
If you prefer to keep the hub focused on your own content only, that is completely fine.
Thanks again for sharing these materials.