Adobe PDF Services AI Agent: Document Data Extraction & Transformation

Unlock Intelligent PDF Automation with this AI Agent

This n8n AI Agent acts as a robust wrapper for the Adobe PDF Services API, enabling intelligent data extraction and manipulation capabilities. It's designed to be triggered with a PDF file (as binary data) and specific Adobe API operation details (e.g., extractpdf for data extraction, splitpdf for splitting pages). The agent handles the authentication handshake with Adobe, uploads your PDF as an asset, initiates the requested processing job, polls for its completion, and finally returns a URL to download the processed result (e.g., extracted JSON data from tables and text, or split PDF files in a ZIP archive).

It leverages Adobe Sensei AI, via the PDF Extract API, for sophisticated document understanding, making it a powerful tool for unlocking valuable information trapped in your PDFs.

Key Features & Benefits

• AI-Powered Data Extraction: Utilizes Adobe Sensei AI for accurate extraction of text, tables, and document structure from PDFs.
• Versatile PDF Operations: Supports various Adobe PDF Services like extractpdf, splitpdf, createpdf, ocr, and more by simply configuring the input endpoint and json_payload.
• Automated Adobe API Interaction: Manages the entire lifecycle: token generation, asset upload, operation initiation, status polling, and result retrieval.
• Streamlined Integration: Designed to be called as a sub-workflow, receiving PDF binary data and operation parameters, then returning the result URL.
• Secure Credential Management: Uses n8n's built-in credential system for securely storing and using your Adobe API keys.
• Status Monitoring: Includes logic to check if the Adobe job is in progress, failed, or successful before attempting to download results.

• An n8n instance (Cloud or self-hosted).
• Adobe PDF Services API credentials (Client ID, Client Secret). Get these by creating a project in the Adobe Developer Console.
• Familiarity with Adobe PDF Services API documentation for understanding available endpoints (e.g., extractpdf, splitpdf, ocr) and their respective JSON payload structures.

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Create two n8n credentials for Adobe API access:
   • Credential 1 (For Token Generation):
     • Type: Custom Auth
     • Name: Adobe API Auth (or similar)
     • Configuration (JSON format):

       {
         "headers": {
           "Content-Type": "application/x-www-form-urlencoded"
         },
         "body": {
           "client_id": "YOUR_ADOBE_CLIENT_ID",
           "client_secret": "YOUR_ADOBE_CLIENT_SECRET"
         }
       }
       

     • Assign this credential to the 'Authenticartion (get token)' node.
   • Credential 2 (For API Calls):
     • Type: Header Auth
     • Name: Adobe API Calls (or similar)
     • Configuration: Field Name: X-API-Key, Value: YOUR_ADOBE_CLIENT_ID (this is the same Client ID used above).
     • Assign this credential to the 'Create Asset', 'Process Query', and 'Try to download the result' nodes.
4. This workflow is designed to be triggered with inputs: endpoint (string, e.g., 'extractpdf'), json_payload (object, specific to the endpoint), and the PDF file as n8n binary data on item.binary.data.
5. For Testing (Manual Trigger Path):
   • The 'When clicking ‘Test workflow’' path is for development.
   • Configure the 'Load a test pdf file' (Dropbox) node with a sample PDF from your Dropbox or replace this node with any other method to load a PDF binary (e.g., Read Binary File).
   • The 'Adobe API Query' node sets a default test endpoint ('extractpdf') and json_payload. Modify these as needed for your specific Adobe PDF Services operation. Consult the Adobe documentation for correct payload structures for different endpoints.
6. The 'Execute Workflow Trigger' node is present if you intend for this workflow to be called by a parent n8n workflow. It expects the PDF binary data on the standard data property of the incoming item from the parent.
7. Review the 'Process Query' node's URL expression to ensure it correctly forms the operation URL: https://pdf-services.adobe.io/operation/{{ $('Query + File + Asset information').item.json.endpoint }}.
8. Activate the workflow. When called, it will return the download URL for the processed file(s) in the final 'Forward response to origin workflow' node.