What is retrieval-augmented generation (RAG)?
Retrieval-augmented generation (RAG) is a method where an AI first looks up relevant documents or data, then writes its answer from what it found. The approach was named in a 2020 paper by Patrick Lewis and colleagues, which paired a language model with a searchable index of documents. Google Cloud describes RAG as combining traditional information retrieval, such as search and databases, with generative language models.
Why it matters to a revenue leader
RAG is how an AI tool answers questions about your business from your own information. Answers about your pricing, agreements, service history or pipeline are only as good as what gets retrieved, so the quality of your documents and CRM data sets the ceiling. It also makes answers checkable, since a well-built RAG system can point to the records it used. When a tool's answer is wrong, the first question is whether it found the right material.
Example
A seller at a fire protection company asks an AI tool what the company charges a customer for annual sprinkler inspections at its three sites. The tool searches the agreement files and the CRM, finds last year's agreement and two change orders, and answers with the current prices and a link to each document. Without retrieval, the model could only guess at typical industry prices.
"When our AI tool answers a question about a customer, can it show which documents or records it used?"