How to deploy Paperless-ngx on Saplo (OCR invoice archive)

Invoices in your email, contracts in a binder, transfer confirmations in the "to send to the accountant" folder - every small business knows this mess. Paperless-ngx is an open-source document archive that scans or accepts files, recognizes text in them via OCR (including Polish), and uploads everything to a searchable database. You type a fragment of an invoice's content, and you get the document in seconds, instead of manually browsing through folders. Below is a step-by-step guide on how to set up Paperless-ngx on Saplo.

Why use Paperless-ngx for invoice archiving

Folders on your disk or in the cloud (Google Drive, Dropbox) work as long as you have only a few dozen documents. With several hundred invoices per year, this method starts to fail:

• searching by file name, not by content - "faktura_2026_03_15_final2.pdf" says nothing about what's inside; you have to open files one by one,
• lack of OCR - for a file search engine, an scanned invoice is just a regular image; you won't find text in it,
• lack of order - categories, tags, and document types must be manually maintained; nobody does this consistently throughout the year,
• lack of history - when a file is overwritten or moved, no one knows who did it or when, which can be a problem during a tax audit.

Paperless-ngx OCRs every uploaded document (Polish language is enabled by default in the Saplo recipe), extracts text from it, and allows you to search by content, date, sender, tag, or document type. It also adds metadata and edit history to each document, so you can see when it was archived and who last changed it. Your entire archive is on your own Box, without paying for extra gigabytes or another account on an external service.

Requirements

To set up Paperless-ngx on Saplo, you need:

• A Box with Docker enabled (the engine automatically enables it when installing from the catalog, container nesting is required, which Saplo has by default),
• A minimum of 2 GB RAM and 15 GB of free disk space on your Box (OCR and full-text indexing are heavier than, for example, Vaultwarden or WordPress, and disk space grows with your scan archive),
• A Domain or subdomain *.saploapp.pl included with every plan.

Paperless-ngx, like Baserow or Cal.com, does not use Saplo's shared database. The recipe sets up its own PostgreSQL container next to the application for document data, and Redis as an OCR task queue broker, so the installation does not require any separate database from the catalog. If your archive grows to several thousand documents per year, it's worth considering the daily backup add-on from the panel, so that a copy of the database and uploaded scans is sent off-Box daily.

Step-by-step installation

1. In the Saplo panel, go to App catalog and select Paperless-ngx (category "Business").
2. Specify the Box on which it will be installed, and the domain or subdomain under which the archive will be available.
3. Confirm the installation. The engine sets up three Docker containers: the official ghcr.io/paperless-ngx/paperless-ngx image, its own PostgreSQL for data, and Redis as an OCR task queue. Then, it exposes the panel under the specified domain via nginx with free SSL.
4. During the first start, the image installs the Polish language pack for OCR and runs database migrations, so the installation takes longer than with lighter recipes from the catalog. Expect several dozen seconds.
5. After completion, the domain responds with the Paperless-ngx login screen.

First login

Unlike Baserow or Cal.com (where the first account is created manually via a registration form), Paperless-ngx gets an administrator account automatically generated during installation, just like Listmonk or Hermes Agent.

1. You can find the administrator login and random password in the PAPERLESS-SETUP.txt file in the application directory on your Box.
2. Log in to the specified domain and immediately change the password in Settings -> Users & Groups.
3. In Settings, you can add more users for other team members or your accountant, each with separate logins and permissions. Access can be restricted so that the accountant only sees documents tagged as invoices, not the entire company archive.

Adding Documents and OCR

1. The simplest way is to drag and drop a PDF file or an image directly into your browser window on the main panel page.
2. Paperless-ngx automatically OCRs the content (Polish language enabled by default in the recipe), recognizes the document date, and suggests a type and tags based on previously set rules.
3. In Tags and Document types, you set your own classification, e.g., "Expense Invoice," "Contract," "Official Letter" - subsequent documents can be automatically matched to these rules without manually tagging each file separately.
4. In the search box, you type any part of the content (contractor name, invoice number, amount) and get the document, even if it was a scan without a text layer.
5. For recurring senders, such as a single electricity provider or a single wholesaler, it's worth setting up rules for automatic type and tag assignment, so that after a few months, most documents classify themselves, and you just confirm the result.

Automation and API Access

Paperless-ngx has a built-in REST API, so you can integrate your archive with automation scenarios instead of handling it manually every time. The most common pattern on Saplo is to set up Paperless-ngx and n8n side-by-side from the same App catalog: n8n receives an attachment from an email inbox or a form, sends it to Paperless-ngx via API along with a pre-set tag and document type, and you get a ready-made archive without clicking in the panel. Similarly, you can connect a synchronization folder (e.g., WebDAV) as a "consumption directory," from which Paperless-ngx automatically downloads new files every few minutes and OCRs them. This is useful when your accountant prefers to upload scans to a shared folder instead of logging into the panel.

Paperless-ngx on Saplo vs. Folders on Disk

FAQ

Does OCR work with Polish characters? Yes, the Saplo recipe enables Polish OCR by default (PAPERLESS_OCR_LANGUAGE and PAPERLESS_OCR_LANGUAGES set to pol), so invoices and contracts in Polish are recognized immediately after installation, without additional configuration.

Where can I find login credentials after installation? In the PAPERLESS-SETUP.txt file in the application directory on Box. The administrator login and password are randomly generated during installation and saved there automatically.

How much disk space does an invoice archive need? It depends on the number and size of scans. The minimum for the recipe is 15 GB. With regular use, it's worth keeping an eye on free space on your Box and adding disk space as the archive grows. A +25 GB SSD add-on can be added from the panel without restarting the application, so the archive doesn't have to be moved to another Box.

Can I add documents by email? Paperless-ngx can receive documents from an email inbox configured in the panel (Settings -> Mail Accounts), so invoices sent by email can go to the archive without manual upload. Simply connect the inbox that receives invoices from suppliers.

Can I have Paperless-ngx and another application on one Box? Yes, every plan has an unlimited number of applications; the only limitations are the RAM and disk of the Box.

What happens to documents if I delete the application? Scans and the database reside in the container volumes on Box, so they remain until the application is manually deleted from the panel. Before deleting, it's worth downloading an archive export (Settings -> Export) or having the backup add-on enabled to avoid losing document history along with the application.

Invoices for laptops and monitors should be linked to the equipment register: IT equipment inventory in Snipe-IT keeps the invoice number, purchase date, and warranty end date for each item.

If you have questions about setting up Paperless-ngx on your own Box or organizing your invoice archive, write to us, we'll help with the initial configuration.