diff --git a/docs-website/docs/concepts/concepts-overview.mdx b/docs-website/docs/concepts/concepts-overview.mdx index a3161eb2c0..b6d37481ea 100644 --- a/docs-website/docs/concepts/concepts-overview.mdx +++ b/docs-website/docs/concepts/concepts-overview.mdx @@ -52,5 +52,17 @@ Finally, you can combine various components, Document Stores, and integrations i If you want to reuse pipelines, you can save them into a convenient format (YAML, TOML, and more) on a disk or share them around using the [serialization](pipelines/serialization.mdx) process. +### Retrieval-Augmented Generation (RAG) + +One of the most common patterns built with Haystack pipelines is Retrieval-Augmented Generation(RAG). In this architecture, a Retriever first finds relevant documents from a Document Store based on a user query. These documents are then provided as context to a Generator powered by a large language model (LLM). + +By grounding the LLM with retrieved documents, RAG systems can produce answers that are more accurate, up-to-date, and traceable to their sources. + +A typical RAG pipeline in haystack consists of: +1. A user query. +2. A Retriever that searches the Document Store. +3. A Generator that produces the final answer using the retrieved context. + + Here is a short Haystack pipeline, illustrated: