Skip to content

The Invoice Fraud Detector Service Initial Release

Latest

Choose a tag to compare

@reory reory released this 26 Feb 11:47
· 10 commits to main since this release

AI Engine: XGBoost model trained with SMOTE to handle imbalanced fraud datasets.

Live Dashboard: Real-time risk speedometer with an SVG-animated needle.

Synthetic Generator: Custom generator.py for creating realistic, mock invoice data.

One-Click Demo: 🎲 "Load Random Sample" feature for instant AI testing.

Developer Ready: Cleanly commented code, .gitignore protection, and full setup guide.

🛠️ Tech Highlights
Backend: Flask API with Pydantic data validation.

Preprocessing: Automated feature engineering (Date/Time extraction & Label Encoding).

Frontend: Vanilla JS/CSS (no heavy frameworks required).