RAG combines a language model with an external knowledge source. Before the model responds, relevant documents are retrieved from a database and embedded in the prompt. This way, the AI answers based on actual sources rather than just training knowledge – which significantly reduces hallucinations and enables current, company-specific answers. RAG is the standard for AI assistants in enterprises.
In short
What does RAG mean in AI?
RAG stands for Retrieval-Augmented Generation. It connects a language model with an external knowledge source: before answering, the AI retrieves relevant documents and relies on them. This reduces hallucinations and enables current, company-specific answers.
Using AI in your organisation?
We show you which applications are worth it for your organisation.