RAGFlow is an open-source platform for building retrieval-augmented generation systems and AI agents around private or enterprise data. It combines document ingestion, parsing, chunking, indexing, retrieval, grounded generation, and workflow capabilities so users can create assistants that answer questions from their own documents with traceable references.

The platform supports heterogeneous files such as documents, spreadsheets, slides, images, and scanned content and provides tools for visually inspecting how information is chunked and retrieved. RAGFlow is useful for knowledge bases, enterprise search, document assistants, and other applications where answer quality depends on reliable retrieval, citations, and control over how source material is processed before it reaches a language model.