ragflow

Why we included this project

RAGFlow is a full retrieval-augmented generation engine that carries you from raw documents to a working pipeline, so teams can get grounded answers from their own data without building the retrieval plumbing themselves. It puts real effort into document understanding, parsing scanned PDFs, tables, and other messy layouts before indexing, and it layers agent capabilities on top of retrieval so you can combine lookups and tool calls into a single answer. The project ships as a self-hostable service with a web UI and REST API, so you can stand it up, point it at your knowledge base, and start querying without wiring together a stack of separate components. Citations and traceable sources come built in, which matters for internal documentation, compliance-heavy content, and any case where an answer is only as good as the evidence behind it.