Invoice and Receipt Extractor

After this module you can turn PDF receipts and invoices into clean spreadsheet rows with AI: no manual data entry.

What it does

Upload a PDF to POST /webhook/receipt. n8n extracts the text, OpenAI pulls out vendor, invoice number, date, currency, subtotal, tax, total, category and line items, the values are cleaned and flagged for review if anything's missing, and the result is appended to an Expenses Google Sheet. Add ?dry_run=true to get the extraction without saving.

Flow

Receipt API → Has PDF? ─true─→ Extract PDF Text → Extract Receipt Data (OpenAI) → Clean Receipt → Dry Run?
                       └false→ 400                                                  ├true─→ Respond
                                                                                     └false→ Save to Expenses → Respond

Verification status

• Upload validation (400s) and PDF text extraction: tested on n8n 2.43.6 with assets/sample-receipt.pdf (the text reaches the AI node, which then stops on the missing OpenAI credential).
• AI extraction: needs an OpenAI credential. WITH_CREDS=1 ./test/test.sh checks 9 fields against the sample receipt.
• Saving: needs Google Sheets. WITH_CREDS=1 SAVE=1 ./test/test.sh.

Files

File	Tier	Purpose
lesson.md	Free	Step-by-step build
workflow/starter.json	Free	Upload + PDF extraction only
workflow/learner.json	Learner	Finished workflow
workflow/business.json	Business	Business workflow (see setup guide)
setup-guide.md	Business	Credentials + configuration
assets/sample-receipt.pdf	Learner	Test receipt with known values
assets/expenses-template.csv	Free	Sheet header row
exercises.md	Learner	Practice tasks
solutions.md	Learner	Exercise solutions
notes.md	Private	Learning log
sell/	Private	Offer, video script, blog post