Your question is RAG Pipeline for Hardware Datasheets. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How would you build a pipeline to ingest, parse, and clean unstructured hardware datasheets so they can be indexed for retrieval-augmented generation (RAG) at Normal Computing?