Inspiration The "Shoebox" nightmare. We realized small business owners aren't lazy; they’re just busy. We wanted to build a tool that accepts financial chaos—crumpled receipts, blurry photos, and handwritten scribbles—and treats it with the precision of a forensic accountant.
What it does Audit Trail turns visual mess into structured data. Users dump photos into the app; Gemini extracts the data into tax-ready JSON, reconciles it against bank feeds to find "The Missing $50," and identifies "Vendor Arbitrage" opportunities where users are overpaying for supplies.
How we built it We leveraged Gemini’s Multimodal capabilities for high-accuracy OCR and handwriting analysis. The backend uses structured output to map raw text to tax categories, while a fuzzy-matching algorithm compares extracted receipt data against bank CSVs for anomaly detection.
Log in or sign up for Devpost to join the conversation.