Julius — AI Finance Bot

Personal finance tracking demands discipline and consistency — two things that manual spreadsheets kill over time. The goal was to build a zero-friction system where receipts go in, structured data comes out automatically, and natural-language queries return spending insights — no manual entry, no app switching, no friction.

Built a fully automated workflow in n8n: receipts are sent via chat, an AI agent (OpenAI) extracts structured data (vendor, amount, category, date) and automatically logs entries into a Google Sheets ledger. Extended Julius with a conversational layer acting as a personal financial manager: users query spending summaries, category breakdowns and budget insights via natural-language chat, with responses grounded in live spreadsheet data. Designed the full workflow architecture including webhook triggers, AI prompt engineering for reliable structured extraction, error handling and dynamic Sheets integration.

A production-grade personal finance automation that runs fully autonomously — zero manual input, real-time ledger updates and conversational financial insights on demand.