https://lesitedefrancois.be/en/tags/embedding/Recent content in Embedding on Le Site de FrançoisHugo -- gohugo.ioen© 2026 Françoishttps://lesitedefrancois.be/en/ai/rag/Mon, 01 Jan 0001 00:00:00 +0000https://lesitedefrancois.be/en/ai/rag/<p><strong>RAG (retrieval-augmented generation)</strong> is an architecture pattern, not a single product: before the <strong>LLM</strong> generates an answer, a <strong>retriever</strong> finds relevant chunks from a knowledge base (wikis, tickets, PDFs, databases) and injects them into the prompt as context. The objective is <strong>grounded</strong> responses—fewer hallucinations on company facts, answers that reflect documents updated yesterday, and traceability to sources—without running full <strong>fine-tuning</strong> every time content changes. A typical pipeline embeds queries and documents with an <strong>embedding model</strong>, stores vectors in a search index, retrieves top-k passages, optionally <strong>reranks</strong> them, then calls the LLM with a system prompt plus retrieved text. RAG is the dominant enterprise pattern for private AI assistants and support bots.</p>