Skip to main content

Extract invoice data from PDFs to JSON with Gemini AI and XML transformation

Workflow preview

Extract invoice data from PDFs to JSON with Gemini AI and XML transformation preview
Open on n8n.io

Important notice

This workflow is provided as-is. Please review and test before using in production.

Overview

This n8n workflow converts invoices in PDF format into a structured, ready-to-use JSON, using AI and XML transformation โ€” without writing any code.

๐Ÿš€ How it works

  1. Upload form โ†’ The user uploads a PDF file.

  2. Text extraction โ†’ The PDF content is extracted as plain text.

  3. XML schema definition โ†’ A standard invoice structure is defined with fields such as:

    • Invoice number
    • Customer and issuer details
    • Items with description, quantity, and price
    • Totals and taxes
    • Bank account details
  4. AI (Gemini) โ†’ The model rewrites the PDF text into a valid XML following the predefined schema.

  5. XML cleanup โ†’ Removes extra tags, line breaks, and unnecessary formatting.

  6. JSON conversion โ†’ The XML is transformed into a clean, structured JSON object, ready for integrations, APIs, or storage.

โœจ Benefits

  • Transforms unstructured PDFs into normalized JSON data.
  • No coding required, only n8n nodes.
  • Scalable to different invoice formats with minimal adjustments.
  • Leverages AI to interpret complex textual content.

๐Ÿ› ๏ธ Use cases

  • Automating invoice data capture.
  • Integration with ERPs, CRMs, or databases.
  • Generating financial reports from PDFs.