Retrieval-augmented generation is a method that supplies an AI model with relevant information retrieved from approved sources before it generates a response.

Why Retrieval-augmented generation matters

It helps agents answer using current product, policy, pricing, or customer information instead of relying only on what the model learned during training.

How it works in practice

A RAG system retrieves a small set of relevant records, passes them into the model's context, and should preserve source boundaries, permissions, and fallback behavior when evidence is weak.