Paperless NGX setup: classify scans without AI

The original video is in German. This tutorial shows how to classify documents in Paperless NGX using scanner presets and filename rules. Daniel Klozbücher presents the method as a more predictable option for documents whose classification controls access rights or later automations. It uses built-in workflows rather than an AI extension.

The setup requires a Paperless NGX installation and access to the scanner's browser management interface. The scanner must support different filenames for scan profiles; the demonstration sends files over FTP. An employment contract preset adds a fixed prefix before the date and counter, then a scanner favorite makes that profile available at the device.

In Paperless NGX, a document-added workflow filters the filename by that prefix, with an asterisk to match the remaining characters. Its assignment action sets the document type and can also set tags, correspondents, or group viewing and editing permissions. The demonstrated scan receives the Employment Contract document type. Klozbücher also explains how workflow placeholders can set a title.

The rule applies to uploaded files with the same prefix. For scanners without filename presets, he suggests email destinations paired with import rules, though this takes more setup. He recommends starting with one or two consequential classifications and retaining ordinary automatic matching for other documents. The workflow relies on choosing the correct preset or filename and configuring the rule and permissions correctly.