Self-hosted Evernote alternative: own your document archive with Paperless-ngx

Paper doesn't stop arriving. Receipts, contracts, tax letters, medical records, warranties, the manual for a dishwasher you'll replace in a decade. Evernote and the apps like it offered to swallow all of it -- then metered your uploads, moved search and PDF handling behind a subscription, and kept your scans on their servers indefinitely. The documents are yours; the archive shouldn't be rented from anyone. Paperless-ngx is the alternative you run yourself: scan a page once, and every letter is OCR'd, tagged, and full-text searchable in a vault that lives where you put it. Then you shred the original.

What Paperless-ngx is

Paperless-ngx is a complete document management system, not a notes app wearing a filing metaphor. You feed it scans, photos and PDFs; it reads the text, sorts them, and gives you a fast searchable archive with a clean web interface. It's the actively maintained community fork of the original Paperless project -- the full-featured tool a small business would otherwise rent as a per-seat SaaS.

The contrast with a hosted service is the whole point. Evernote and its peers put document OCR, PDF search, larger uploads and Office-file handling behind their paid tiers, and every scan you add deepens the lock-in. Paperless-ngx hands you all of that with no plan to upgrade to and no upload ceiling but your own disk. Your documents sit in a database and a folder you control, not a vendor's cloud you're renting month to month.

The features that actually matter

• OCR that makes paper searchable. Every incoming scan or image runs through Tesseract OCR, which adds an invisible, selectable text layer -- so a photographed receipt becomes text you can search and copy. Over 100 languages are supported. On Evernote this quality of PDF and document OCR is a paid-tier feature.
• Full-text search that actually finds things. The whole archive is indexed, with fuzzy matching and highlighted results. Type "roof warranty 2019" and it finds the letter even if that's the only place those words appear.
• Automatic tagging and sorting. Paperless-ngx trains a classifier on the documents you've already filed, then auto-assigns tags, correspondents and document types to new arrivals. The more you file, the less you touch.
• Matching rules you control. Beyond the machine learning, you can define correspondents, document types and storage paths with literal, regex or auto-matching -- anything from "Blue Cross" to a pattern that recognises your utility bills.
• Import from anywhere. A watched consume folder ingests whatever your scanner drops into it, and the built-in email fetcher pulls attachments straight from an IMAP inbox. Documents process themselves with nothing to click.
• Barcodes and archive serial numbers. It can split a single multi-document scan on barcode separators and stamp each with an ASN, letting you tie a digital copy back to a physical box if you keep originals.
• Custom fields, workflows and PDF/A archiving. Add your own metadata fields, run workflows that act on documents as they arrive, and keep both the untouched original and a searchable PDF/A archive copy of everything.

Why self-host it instead of the hosted version

Three reasons, and the first is the honest one. Your documents are as private as data gets. Tax records, medical letters, contracts, IDs -- a self-hosted archive means no third party stores, scans or profiles any of it. The scans never leave infrastructure you control, and nobody's terms of service get a vote.

The cost doesn't scale per seat. Hosted document apps charge per user, per month, forever, and nudge you upward as your library grows. Paperless-ngx is free software; you pay for the box it runs on and nothing else, whether you file ten documents a month or ten thousand, alone or with your whole household.

You keep full control. It's your database, your storage folder, your retention rules, your backup schedule. Your originals are stored as ordinary files, and a documented REST API means the archive is scriptable and exportable rather than trapped in a proprietary format. You decide who gets an account and what they can see, you export the lot whenever you want, and no feature you rely on can be moved behind a new paywall next quarter.

Who it's for

Households drowning in paper -- the family that wants tax season to take an afternoon instead of a weekend, with every receipt, statement and warranty a search away.

Small businesses and sole traders who need contracts, invoices and correspondence organised and findable, but don't want a per-seat DMS subscription or their books sitting in someone else's cloud.

Privacy-minded people and professionals -- anyone handling medical, legal or financial documents who simply shouldn't be uploading them to a consumer notes service, and would rather own the archive outright.

The honest setup notes

We'll be straight with you: Paperless-ngx self-hosted is powerful and it is a bit of a chore. It isn't one container -- it needs a PostgreSQL database and a Redis broker alongside the app, plus optional helpers if you want to ingest Word and Excel files. OCR is genuinely CPU-hungry, so a large import can peg a small server for a while, and the whole thing wants a sensible amount of RAM and disk headroom to stay comfortable.

Then there's the ongoing part. This is your document archive, so backups are not optional -- you're protecting the database and the media folder together, and a snapshot you never tested isn't a backup. Updates land regularly and occasionally touch the database. If you want email import or notifications you'll wire up IMAP and SMTP yourself.

None of it is exotic, but it's real ops work. Or let us run it, and the chores become ours.

Deploy it in one click

Rolling your own means standing up the database, the broker, the OCR pipeline and the optional file converters, getting the volumes and permissions right, then owning the backups, the updates and the TLS certificate from there on. It's doable -- people do it -- but it's an afternoon before you've filed a single page.

On Caliber Node it's one click. We provision the full stack, wire up the database and broker, put it behind HTTPS on a domain, and keep the pieces patched and backed up. You get a running Paperless-ngx and a login, not a compose file and a weekend. When you're ready, deploy Paperless-ngx in one click and start dropping documents in minutes instead of hours.

First run

1. Log in to your new Paperless-ngx instance and create your admin account.
2. Upload your first document -- drag a PDF or a phone photo of a receipt straight into the web interface, or drop it into the consume folder.
3. Watch it process: Paperless-ngx OCRs the page, builds a searchable archive copy, and lands the document in your inbox view.
4. Give it a tag, a correspondent and a document type -- or let the auto-matcher suggest them once it's seen a few examples.
5. Search for a word you know is inside it. When the document surfaces in a second, your archive is working -- now feed it the pile.

The bigger picture

A lifetime of documents is not a thing to rent. Every receipt and contract you file is a small bet that the tool holding them will still be there, still yours, still affordable in ten years -- and a hosted notes service is a shaky bet on all three. Owning this category means the archive outlives any company's pricing change, and the answer to "where's that letter?" is always a search box you control. That's worth the modest effort of running it -- or the single click of letting us.

Deploy Paperless-ngx in a few minutes → https://calibernode.com/cloud-apps

Or browse the full app catalog to see what else you can own.

One-click self-hosted apps, managed for you -- real logs, SFTP, and your data always exportable. Never a black box.