Automating Invoice Data Extraction with OCR and LLM Workflow

steps

1. Create an API Key in Google AI Studio:

   • Log in to Google AI Studio.
   • Navigate to the "API Keys" section and generate a new key.
   • Save the key for use in your workflow.

2. Upload a Sample Invoice Image:

   • Prepare PDFs or image files of invoices and upload them to your source folder (e.g., Google Drive).

3. Setup OCR Model Configuration in n8n:

   • Add an HTTP request node to n8n for interacting with Gemini 2.5 Pro.
   javascriptCopy

   const options = {
       method: 'POST',
       url: 'https://api.gemini.com/v1/ocr',
       headers: { Authorization: `Bearer YOUR_API_KEY` },
       data: {
           file: '<encoded_image_base64>',
           options: { language: "en", outputFormat: "json" }
       }
   };

4. Process OCR Outputs Using an LLM:

   • Include a node to parse the OCR output and extract desired fields (e.g., date, amount).
   javascriptCopy

   const extractedData = {
       date: "2023-10-15",
       amount: "123.45",
       paymentMethod: "Credit Card"
   };

5. Store Extracted Information in Google Sheets:

   • Use n8n to send parsed data to a row in Google Sheets.
   javascriptCopy

   const sheetData = [
       ["Invoice Date", "Vendor", "Amount"],
       ["2023-10-15", "ABC Supplies", "123.45"]
   ];

• Test the full workflow for multiple files to ensure consistent operation.