n8n Google Drive OCR & GPT-4o Extract to Google Sheets

Automatically OCR documents in Google Drive and log structured case data to Google Sheets

This n8n workflow watches a Google Drive inbox folder, downloads newly added files, uses OpenAI GPT-4o-mini to OCR and extract strict structured metadata, then appends the results to Google Sheets—archiving successful files and routing failures to an error folder.

What this workflow does

• Triggers on new documents in a specified Google Drive inbox folder (or can process all existing inbox files when run manually).
• Reads existing Google Sheets rows to provide context for case assignment and duplicate detection.
• Downloads each Drive file and sends it to OpenAI (GPT-4o-mini) to perform OCR and return a strict JSON extraction.
• Extracts structured fields such as sender, type, dates, deadlines, amounts, tax relevance, case linkage, and duplicate information.
• Normalizes output into a Google Sheets row, generating a new case ID and document ID when needed, enforcing consistency rules (including clearing the “next action” when not required), and setting a status based on any extracted deadline.
• Appends to Google Sheets, waits briefly to reduce rate-limit pressure, then moves the processed file to a Drive archive folder.
• Error handling: if AI analysis fails, it logs an error row in Google Sheets and moves the file to a dedicated Drive error folder.

Use cases

• Capture and classify invoices or correspondence dropped into a Google Drive inbox.
• Automatically extract deadlines and amounts, then update case tracking in Google Sheets.
• Prevent duplicate records by comparing AI-extracted metadata against existing Sheets rows.

Technical details

• n8n workflow with: Manual Trigger, Code, Google Drive, Google Sheets, No Op, and Wait.
• Uses OpenAI GPT-4o-mini for OCR + strict JSON extraction.
• Requires Google OAuth2 credentials and configuration of inbox, archive, and error folder IDs in Google Drive, plus the target Google Sheets document.