Paperless Extension¶

Manage documents with Paperless-ngx, an open-source document management system with powerful OCR and full-text search.

Why Paperless-ngx?¶

What You Can Do¶

Search Documents¶

Full-text search:

Find documents by their content thanks to OCR.

Filter by metadata:

Narrow results by correspondent, tags, or document type.

Find unprocessed:

See documents that haven't been classified yet.

Example requests:

• "Find invoices from 2024"
• "Search for documents containing 'contract renewal'"
• "Show unclassified documents"
• "Find documents tagged as 'Urgent'"

Read Documents¶

Get OCR text:

Read the extracted text from any document.

View metadata:

See tags, correspondent, document type, and dates.

Find similar:

Discover related documents automatically.

Example requests:

• "Read the content of document 42"
• "Show details for the latest invoice"
• "Find documents similar to this one"

Add Documents¶

Upload files:

Add new documents to Paperless.

Set metadata:

Assign title, correspondent, and tags on upload.

Monitor processing:

Check if OCR is complete.

Example requests:

• "Upload this PDF to Paperless"
• "Add this invoice from Supplier GmbH with tag 'Unpaid'"

Organize Documents¶

Update metadata:

Change tags, correspondent, or document type.

Bulk operations:

Apply changes to multiple documents at once.

Auto-tagging:

Set up automatic classification rules.

Example requests:

• "Add the 'Reviewed' tag to document 42"
• "Change the correspondent to 'New Supplier GmbH'"
• "Tag all invoices from last month as 'Q4-2024'"

Key Concepts¶

DeskAgent Auto-Tagging¶

DeskAgent includes a powerful Auto-Tagging Agent that uses AI to automatically classify your documents. This is especially useful for processing months of unclassified documents.

How It Works¶

graph LR
    A[Scan unclassified docs] --> B[Read OCR text]
    B --> C[AI analyzes content]
    C --> D[Suggest tags + correspondent]
    D --> E[You confirm]
    E --> F[Batch classify]

1. Find unclassified: DeskAgent searches for documents without a correspondent
2. Read content: AI reads the OCR text of each document
3. Classify: AI determines the document type, sender, and appropriate tags
4. Confirm: You review the suggestions before applying
5. Apply: All changes are made in one batch operation

Recommended Tag System¶

Set up these tags in Paperless for a complete filing system:

Using Correspondents as Status Marker¶

Pro tip: Use the correspondent field to track classification status:

• No correspondent = Document needs classification
• Has correspondent = Already processed

This makes it easy to find unprocessed documents:

"Show all documents without correspondent from last month"

Storage Paths for Business vs Private¶

Organize documents by ownership:

DeskAgent can automatically set the storage path based on document content.

Example: Monthly Classification¶

Say: "Classify all untagged documents from January 2025"

DeskAgent will:

1. Find all documents from January without a correspondent
2. Read each document's OCR text
3. Show you a table with suggested classifications:

1. After your confirmation, apply all changes at once

Workflow Examples¶

Single Invoice Processing¶

1. Upload: "Upload the invoice PDF from my Downloads"
2. Wait for OCR: Paperless processes the document (usually seconds)
3. Review: "Read the content of the new document"
4. Classify: "Set correspondent to Supplier GmbH and tag as Invoice"
5. Later: "Mark all invoices from last month as 'Paid'"

Bulk Classification¶

1. Find: "Show unclassified documents from December"
2. Auto-tag: "Classify these documents automatically"
3. Review: Check the AI suggestions
4. Apply: Confirm to apply all changes

Tax Preparation¶

1. Export: "Find all documents tagged 'AnSteuerberater' from 2024"
2. Download: "Download these as a ZIP file"
3. Send: Ready for your accountant

Matching Rules¶

Paperless can also classify documents automatically using built-in rules:

Setup¶

Step 1: Get Your API Token¶

In Paperless-ngx:

1. Go to Settings > Users
2. Click on your user
3. Find or generate an API token

Or ask DeskAgent: "Get Paperless token for username password"

Step 2: Configure DeskAgent¶

1. Open DeskAgent Settings
2. Go to Integrations
3. Find Paperless
4. Enter:
5. Server URL (e.g., http://localhost:8000)
6. API Token
7. Click Test Connection

Paperless vs ecoDMS¶

Tips¶

1. Let OCR complete: Wait a few seconds after upload before reading content.

2. Use correspondents: Set up correspondents for frequent senders.

3. Tag system: Create a consistent tagging system (e.g., Year, Status, Type).

4. Bulk classify: Update multiple documents at once for efficiency.

Common Issues¶