clusterzx/paperless-ai

Paperless-AI is an add-on for Paperless-ngx, an open source document management system that scans and organizes physical and digital documents. This add-on connects Paperless-ngx to AI language models so that incoming documents are analyzed and categorized automatically. When a new document arrives, the system reads its content and assigns a title, tags, document type, and correspondent without the user having to do it manually. The AI backends it can connect to include OpenAI, Ollama (which lets you run AI models locally on your own hardware), DeepSeek, Google Gemini, and several other compatible services. Running models locally through Ollama means documents never leave your machine, which the README notes as a privacy benefit. Beyond automatic tagging, the tool includes a chat feature powered by a technique called Retrieval-Augmented Generation (RAG). RAG is an approach where the AI searches through your actual documents to answer questions, rather than relying purely on its training. You can ask things like "when did I sign my rental agreement" or "which documents mention my health insurance" and get answers drawn from the specific files in your archive. There is also a manual processing mode, accessible through a web interface, for reviewing and tagging sensitive documents yourself rather than having AI handle them automatically. Tag rules let you control which documents get processed and what output tags are applied. The project is distributed as a Docker container, which is a packaged format that simplifies installation. The README includes an important notice that the project is not currently being actively maintained, and the author is considering whether to continue depending on how official AI support in Paperless-ngx itself develops.