Paperless-ngx
Scan, OCR and search every paper document you own
Paperless-ngx turns scans and PDFs into a searchable archive: it runs OCR, guesses the correspondent, type and tags, and files everything for you. It is GPL-3.0, community maintained, and a common first app for people drowning in paperwork.
Best for when
- you want every scanned letter and invoice searchable
- you want tagging and filing to happen automatically
- you want sensitive papers on your own hardware
Watch out for
- it holds some of your most sensitive documents, so private access and encrypted backups are essential
- OCR quality depends on scan quality and the languages you configure
- automatic matching needs a few weeks of corrections before it gets things right
Paperwork you can actually find
Drop a scan into a folder, email it in or upload it, and Paperless-ngx reads the text, stores an archived PDF, and applies tags and dates based on rules it learns from your corrections. Finding a tax letter from three years ago becomes a search instead of an afternoon.
It fits well with a phone scanning app and a habit: scan on arrival, let it file, throw away the paper you do not legally need. The full text search is the part people miss once they have had it.
What it takes
The stack is the web app, a task worker, Redis and usually PostgreSQL, with OCR being the CPU heavy part. Storage is small compared with photos or video, but these are some of your most sensitive files.
That sensitivity sets the rules: keep it private, encrypt the backups, and use its built in exporter so you have a copy that does not depend on the app. A document archive with no backup is just a slower way to lose papers.
