Paperless-ngx

Paperless-ngx runs OCR (Optical Character Recognition) on every document, extracting the text so that every word becomes searchable from the web UI. It then applies machine learning to the content to automatically assign tags, correspondents, and document types. The result is that you find a receipt or contract by full-text search instead of digging through folders.

Paperless-ngx is designed to run in Docker on hardware you already own — a Raspberry Pi or small home server is more than enough for a personal archive. OCR is the main CPU-heavy task, so larger libraries or high-resolution scans benefit from more processing power. Backups are straightforward because the archive lives on disk.

You can upload documents through the web UI, but Paperless-ngx also watches a folder on your server, can consume documents delivered to a dedicated email address, and integrates with common document scanners. Anything that lands in these inputs is OCR'd, classified, tagged, and filed automatically. That enables a scan-and-forget workflow where documents organise themselves.