Document.ai is an open-source local knowledge-base solution that lets you chat with your own documents using a vector database and GPT-3.5. Built early in the LLM wave, it offered a universal recipe for private document Q&A: drop in your files, embed them into a vector store, and ask questions in natural language with answers grounded in your material. The project has around 3,700 stars on GitHub and is released under the AGPL-3.0 license.
The system splits uploaded documents into chunks, embeds them with an OpenAI embedding model and stores the vectors locally, then retrieves the most relevant passages at question time and feeds them to GPT-3.5 as context. Because the knowledge base lives on your own machine, it suits personal notes, internal manuals and research collections that you do not want to upload to a third-party service. The project supports common document formats and was designed to be simple to deploy for individuals and small teams experimenting with retrieval-augmented generation before dedicated RAG frameworks became commonplace.
Document.ai is completely free and open source under the AGPL-3.0 license. You only pay for the OpenAI API usage for embeddings and chat completions.
Individuals and small teams who want a self-hosted way to question their own document collections, and developers learning the fundamentals of retrieval-augmented generation from a compact, end-to-end codebase.