Tool calling and retrieval (RAG)
Give models tools and knowledge.
- AGT.06.01Tool callingCoreMid-level10 min
- AGT.06.02Retrieval basics: chunking, embeddings and vector searchCoreMid-levelBeing written
- AGT.06.03Better retrieval: hybrid search, reranking and citationsCoreMid-levelBeing written
- AGT.06.04Evaluating RAGCoreMid-levelBeing written
- AGT.06.05Document parsing and ingestion for RAGMid-levelBeing written
Interview questions in this module
- Explain how tool calling works end to end.
- How do you design good tools for an LLM?
- Walk through a RAG pipeline. Where does it usually fail?
- How do you choose a chunk size?
- What is hybrid search and why does reranking help?
- How do you evaluate a RAG system?
- When would you skip RAG and use a long context window instead?