Your question is Evaluate RAG Retrieved Context. 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).
You are building a system where a language model answers questions from a fixed document collection instead of relying only on its pretrained knowledge. You have a Retrieval-Augmented Generation pipeline in place, and you want to judge whether the retrieved passages actually support the final answer.
Explain the architecture of a RAG system and how you would evaluate the quality of its retrieved context.