paperless-ai

Why we included this project

Anyone running Paperless-ngx knows the manual part of going paperless: every scanned invoice and contract needs a title, tags, and a correspondent before it's actually filed. Paperless-AI watches that document store for new files, runs their contents through an OpenAI-compatible API or a local Ollama model, and fills in those fields for you. It also comes with a retrieval-augmented chat that answers questions like "which documents mention my health insurance?" using your own archive's text instead of keyword matching, so you can find things without guessing at tags. That combination is genuinely useful for individuals and small offices, though it's worth knowing the author has paused maintenance while rewriting the codebase, so plan to deploy or fork it as a starting point rather than rely on ongoing updates.