AI Glossary · Definition
What is Retrieval-augmented generation (RAG)?
Retrieval-augmented generation (RAG): RAG is a technique where an AI system first retrieves relevant passages from your own documents and then uses them to generate an answer, ideally with links to the sources.
By DAIDU EditorialUpdated
Retrieval-augmented generation (RAG), explained
LLMs only know what they learned in training, and they do not know your company's policies or products. RAG solves this by searching a prepared collection of your documents for passages related to the question, adding those passages to the prompt, and asking the model to answer only from them. This reduces made-up answers and lets you update knowledge by updating documents rather than retraining a model. Quality depends on clean, current documents and on good retrieval.
Example
An HR assistant answers 'How many days of annual leave do I get?' by retrieving the leave policy section and quoting it with a link.