Your question is End-to-End RAG for Support. 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).
Design an end-to-end RAG pipeline for a customer support assistant, including retrieval, prompt construction, and response generation.
Explain how you would ingest and chunk support content, retrieve and rerank relevant passages, construct a grounded prompt, and generate a safe response. Include evaluation for retrieval quality, answer faithfulness, hallucination, prompt injection, latency, and cost.