Automate OCR and Document Type Detection for Scanned PDFs

Step 1: Monitor for New Scanned Files

Node: Google Drive Trigger

Settings:

• Trigger On: Changes Involving a Specific Folder
• Folder from List: Select your "incoming" or "scanned documents" folder
• Watch For: File Created

Success Looks Like: The trigger activates immediately when a new scanned file is added to your watched folder.

Step 2: Perform OCR on Scanned Document

Node: PDF.co API → Make PDF Searchable or Unsearchable

Settings:

• URL: ={{ $json.webContentLink }}
• Operation: Make PDF Searchable
• Language: English (Adjust for your documents)
• Advanced Options:
  • File Name: ={{ $json.originalFilename }}

What This Does: Applies OCR to scanned, image-based PDFs and adds an invisible text layer, making the document fully searchable and selectable without altering its visual appearance.

Success Looks Like: PDF.co processes the document and returns a URL to the searchable version. The response includes the processed file URL that can be used for classification.

Step 3: Classify the OCR-Processed Document

Node: HTTP Request → PDF.co Document Classifier

Settings:

• Method: POST
• URL: https://api.pdf.co/v1/pdf/classifier
• Headers:
  • x-api-key: Your PDF.co API key
  • Content-Type: application/json
• Body (JSON):

{
  "url": "{{ $json.url }}",
  "rulescsv": "Invoice, OR, Invoice Number, Invoice #, INVOICE NO\nPurchase Order, OR, PO Number, Order Number, Order No\nBill, OR, Bill Date, Billing Period, Bill Number"
}

Customize Your Classification Rules

The rulescsv parameter defines how documents are classified. Format: FolderName, OR, keyword1, keyword2, keyword3

Examples:

• Invoice, OR, Invoice Number, Invoice #, INVOICE NO
• Purchase Order, OR, PO Number, Order Number, Purchase Order No
• Receipt, OR, Receipt, Transaction, Purchase Date
• Contract, OR, Agreement, Contract, Terms and Conditions
• Statement, OR, Statement, Account Statement, Monthly Statement

Success Looks Like: PDF.co analyzes the OCR'd document content and returns a classification result with the matched category name. The response includes a classes array with the best match.

Step 4: Find the Destination Folder

Node: Google Drive → Search files and folders

Settings:

• Resource: File/Folder
• Operation: Search
• Search Query: ={{ $json.body.classes[0].class }}
• Filter:
  • Folder From List: Your main organizational folder (parent of all category folders)
  • What to Search: Folders
• Return All: Yes

Success Looks Like: The search returns the folder that matches the classification result. If "Invoice" was classified, it finds your "Invoice" folder.

Step 5: Download the Processed File

Node: HTTP Request (Download OCR'd File)

Settings:

• Method: GET
• URL: ={{ $('PDFco Api').item.json.url }}

What This Does: Downloads the OCR-processed, searchable PDF file so it can be uploaded to the destination folder.

Success Looks Like: The processed file is downloaded and ready for upload to the classified folder.

Step 6: Upload File to Correct Folder

Node: Google Drive → Upload file

Settings:

• Operation: Upload
• Input Data Field Name: data
• Parent Drive from List: My Drive
• Parent Folder by ID: ={{ $json.id }} (from the search result in Step 4)

Success Looks Like: The processed, searchable file appears in the correctly classified destination folder.