Paperless-ngx is the project behind this setup. Start with disposable sample documents. The first successful test is not OCR—it is restoring the archive after pretending the server failed.
- Recommended method
- Official Docker Compose setup
- Difficulty
- Intermediate
- Typical time
- 45–75 minutes
What you need
- A Docker host
- Persistent storage
- The correct OCR language configuration
- A protected backup destination
Set it up
Use the official setup generator
Paperless-ngx provides an interactive Docker Compose setup. Use it instead of copying an old compose file.
Choose storage and OCR settings
Set the timezone, OCR languages and persistent paths before the first import.
Start the services and create an administrator
Confirm the web app is reachable only from the network or secure access path you intended.
Import three sample documents
Use different layouts and confirm OCR text, dates, correspondents and document types can be corrected.
Export and restore
Follow the official administration guidance and restore into a temporary instance before adding real records.
Reserved for launch
Verify it works
Do not call the setup finished until these small tests pass:
- OCR finds text inside a scanned PDF
- Search retrieves the document
- Original files remain available
- Metadata survives export and restore
Protect the service
- Keep identity and financial records off the test instance
- Use individual accounts and least privilege
- Do not expose the service without HTTPS and access controls
- Update promptly when security releases appear
Official resources
Videos are supplemental. Use the official documentation for current commands, versions and security guidance.