Your question is RAG With 200ms Context Fetch. 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 implement a retrieval-augmented generation (RAG) pipeline that can fetch relevant customer context within a 200-millisecond window during a call for Superdial?