Retrieval-Augmented Generation (RAG)

Retrieval-augmented generation (RAG) is a technique where an AI searches a specific document store first, then writes its answer only from what it retrieved. The output is grounded in real sources instead of the model's memory — which is where made-up answers come from.

It's the difference between an assistant that quotes and one that guesses. Gatherly's knowledge base is built on RAG: semantic search across your documents and Drive files, answers with citations you can check in one click. Hold any AI chief of staff to that standard — a briefing or meeting prep note is only as good as the sources behind it.