Automating Financial Report Extraction with n8n, PDF.co, and Google Drive

Step 1: Trigger on Financial Reports

Node: Email Trigger (IMAP)

Settings:

• Mailbox Name: INBOX
• Action: Mark as Read
• Download Attachments: True

Custom Email Rules: ["UNSEEN", ["SUBJECT", "financial_report"]]

Success looks like: When a new/unread email with subject financial_report arrives, the workflow is triggered and the attached PDF is pulled in.

Step 2: Upload File to PDF.co

Node: PDF.co API → Upload File

Settings:

• Operation: Upload File to PDF.co
• Binary Data: True
• Binary Property Name: file0
• File Name: ={{ $json.subject }}

Success looks like: The PDF financial report is uploaded to PDF.co and returns a secure file URL.

Step 3: Search for Invoices

Node: PDF.co API → Search in PDF

Settings:

• URL: ={{ $json.url }} (from Step 2)
• Search String: "Invoice No"

Success looks like: PDF.co returns an array of all pages containing the text Invoice No.

Step 4: Extract Page Numbers

Node: Code (JavaScript)

// Get the body array from the previous node
const body = $input.first().json.body || [];
// Extract pageIndex values
const pageIndexes = body.map(entry => entry.pageIndex);
// Convert to human-readable page numbers (index + 1), remove duplicates, and sort
const pageNumbers = [...new Set(pageIndexes.map(i => i + 1))].sort((a, b) => a - b);
// Return result
return [{
  json: {
    pageNumbers: pageNumbers,             // [1, 4, 5]
    pageNumbersCsv: pageNumbers.join(",") // "1,4,5" for APIs that need string
  }
}];

Logic:

• Takes the array of pageIndex values from Step 3.
• Converts them into human-readable page numbers (index + 1).
• Removes duplicates and sorts them.

Exports results as:

{

"pageNumbers": [1, 4, 5],

"pageNumbersCsv": "1,4,5"

}

Success looks like: Invoice page numbers are extracted and formatted for the Split API.

Step 5: Split PDF to Extract Invoices

Node: PDF.co API → Split PDF

Settings:

• URL: ={{ $('Assign URLs to Binary Files').item.json.url }}
• Pages: ={{ $json.pageNumbersCsv }}

Success looks like: A new set of PDF files, each containing only invoice pages, is returned.

Step 6: Parse Customer’s Name

Node: PDF.co API → Parse

Settings:

• URL: ={{ $json.body[0] }}

Success looks like: Customer information (e.g., company name, contact details) is extracted from the invoice.

Step 7: Merge Files

Node: PDF.co API → Merge PDF

Settings:

Input URLs: 

={{ $('Split File to get Invoices').item.json.body[0] }}

={{ $('Split File to get Invoices').item.json.body[1] }}

Output File Name: {{$json.body.vendor.name}}_merged_invoice.pdf

Success looks like: Multiple invoices are combined into a single professional PDF.

Step 8: Convert to Binary File

Node: HTTP Request

Settings:

• URL: ={{ $json.url }}
• Batch Size: 1
• Batch Interval: 500 ms

Success looks like: The merged invoice is converted into binary format for upload.

Step 9: Upload to Google Drive

Node: Google Drive

Settings:

• Drive: My Drive
• Folder: OUTPUT1 (Google Drive folder link)

Success looks like: The final invoice PDF is saved to your Google Drive folder.

Step 10: Final End-to-End Test

1. Send an email with subject financial_report and a PDF attachment.
2. Execute the workflow in n8n.
3. Check your Google Drive folder → the extracted and merged invoices should be saved.