Skip to content
ML Academy

RAG and agents

Interview question

  • midlevel
  • 8 minsuggested time
  • 2Applied AI, ML engineer

Walk me through the components of a retrieval-augmented generation (RAG) system, and name one way each component typically fails.

Tests: End-to-end understanding of a RAG pipeline: ingestion, chunking, embedding and indexing, retrieval, reranking, context assembly, generation with citations; component-level failure modes.

day 47day 48day 51

Sign in to practise: answers are saved before model answers are shown.