A fully automated workflow built in n8n that extracts structured financial data from uploaded receipts and invoices using AI, validates and formats the output, and logs all entries into Google Sheets for bookkeeping and property-management tracking.
This project demonstrates my ability to design and build real-world AI automation solutions using n8n, OpenAI, and Google Workspace.
The workflow accepts a document upload from a custom form, processes it with an AI extraction model, and returns structured fields including vendor, date, amount, category, and payment type. The automation then sanitizes the extracted values, checks for missing data, and handles fallback logic if the AI returns incomplete results.
Once validated, the flow writes a new row into a Google Sheet, serving as a simple accounting ledger for property management. Future versions include auto-categorization, OCR fallback for low-quality scans, and a mobile app for snapping photos of receipts on the go.
This system replaces inconsistent manual entry with a clean automation pipeline—saving time, reducing errors, and giving property managers real-time visibility into expenses.