Back to All Posts
Jun 9, 2026 3:27:39 PM

What is RAG (Retrieval-Augmented Generation)?

By dotsquares 1 minute read

RAG combines a retrieval system (vector database + embeddings) with an LLM. At query time, relevant documents are fetched and injected into the prompt, grounding the model’s response in up-to-date or proprietary data without retraining. It is the dominant pattern for knowledge-intensive applications.

Add Comment







    ↻