RAG

Retrieval-Augmented Generation

In one sentenceA way of making AI answer from your documents: it looks up the relevant information first, then writes the answer from it.

What it means

A model only knows its Training Data. RAG adds a search step: when you ask a question, the system retrieves the most relevant passages from a knowledge base (your docs, website, help center, CRM notes), places them in the prompt, and the model answers from them.

This is how most "chat with your documents" tools and company help bots work. It reduces hallucinations and keeps answers current without retraining the model.

How to use it

• The simplest RAG: upload your files to a Claude Project, a custom GPT or a NotebookLM notebook and ask questions about them.
• For a website help bot, feed it your FAQs, pricing and policies so it answers from those, not from guesses.
• Garbage in, garbage out: clean up outdated documents before you load them.

Related terms

Embeddings Vector Database Knowledge Base Grounding Semantic Search