RAG: AI over your own documents

How tools answer from your files instead of from memory.

What it is

Retrieval-augmented generation finds the passages in your documents most relevant to a question, then gives only those to the model to answer from. It's how NotebookLM and “chat with your PDF” tools work.

Why it matters

Grounding answers in known sources cuts down made-up facts and makes answers checkable.

Key ideas

• Documents are split into chunks and indexed by meaning (embeddings).
• Bad retrieval means bad answers, however good the model is.
• Always check the cited passage.

Try it: Stress-test a grounded answer

1. Load three readings into NotebookLM.
2. Ask a question the readings don't answer. Does it say so?
3. Ask one they do answer, and check the citation.