Set up invoice capture and bookkeeping in n8n

This n8n workflow pulls fields from uploaded invoices or receipts, lets the user check them, then stores the file and logs the data in a yearly sheet.

Done with n8n· Intermediate

What they did
The workflow started with a web upload that sent an image or PDF to an n8n webhook. It matched the uploader against a token stored in Google Sheets, then passed pictures to OpenAI vision and pulled text from PDFs before analysis. The model produced structured JSON, including invoice date, merchant, gross total, VAT rate, and category, and the user could edit the result before saving.
How it went
Once approved, the original file was moved into a Drive folder arranged by user and month, and the extracted fields were added to a yearly spreadsheet. The public project was shown as a working prototype with a demo video, but the exported files were sanitized and still needed the user’s own setup before they would run.
Tips
Plan to connect your own OpenAI and Google OAuth accounts, plus a sheet for user lookup, a root Drive folder, and an n8n webhook address. The token check is only a stand-in for proper authentication, so keep it as a prototype pattern.

Try it yourself with n8n

Create an n8n automation that accepts a [receipt or invoice] from [upload page or web form], identifies the sender with [token or login check], and pulls out [date], [merchant], [gross amount], [VAT rate], [category], and any other fields that help with bookkeeping. If the file is a PDF, extract its text first; if it is an image, use vision analysis. Present the parsed results for review and pause there until I approve. After approval, store the original in [Drive folder structure] and add the verified row to my [year] bookkeeping sheet.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

2 more reports of this

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.