I automatically track online spending (invoices) in a spreadsheet (blog.terrydjony.com)

A Google Apps Script add-on lets you automatically extract invoices from Gmail and write them into a Google Sheets expense tracker (or download as CSV). The sidebar UI asks for a start/end date and your OpenAI API key (stored locally), then scans matching emails, uses the OpenAI API to parse invoice content and currencies, and outputs structured rows into the active sheet. The tool supports multi-currency detection and conversion-friendly formatting, and exposes a simple “Process Invoices” → “Insert to Sheets / Download CSV” workflow. For the AI/ML community this is a practical example of embedding LLM-based extraction into everyday automation: it shows how lightweight orchestration (Google Apps Script + LLM) can turn unstructured email receipts into structured financial records. Key technical implications include reliance on the OpenAI API (so consider API costs, rate limits, and prompt/response design), Gmail scopes and data-privacy concerns (email content is sent to an external model unless you run a locally hosted alternative), and the opportunity to extend the parser for categories, tax fields, or accounting integrations. It’s a useful reference implementation for anyone building LLM-powered data pipelines or personal productivity tools that convert text-heavy artifacts into analyzable datasets.