Back to problems

RAG Q&A Chatbot — ML / AI Technical Deep Dive

System Design · Vanta · Hard

Requirements Propose and critically assess a question-answering chatbot built on retrieval-augmented generation (RAG). Explain how the system should ingest and index both structured and unstructured data so the retrieval component can make use of them. Lay out how you would judge performance: metrics for answer quality and relevance, plus an evaluation methodology. For a set of failure scenarios, describe your debugging approach and how you would rank the improvements to…

Checking your access…