A full-stack web application that helps retail stores manage inventory, analyze sales trends, forecast product demand using machine learning, and optimize restocking decisions.
- Built a full-stack AI-powered retail management system using Django REST Framework and React (Vite), featuring JWT authentication, paginated REST APIs, and a responsive dashboard UI
- Implemented demand forecasting using Scikit-learn's Random Forest Regressor and Linear Regression with lag features, rolling averages, and seasonal signals; evaluated with MAE and R² metrics
- Designed a normalized relational database (7+ tables) with proper indexing and foreign key relationships for inventory, orders, and sales tracking
- Built an AI restock recommendation engine that calculates per-product daily sales velocity and estimates days-to-stockout to prioritize critical restocking actions
- Developed a co-purchase product recommendation system using market basket analysis to surface frequently bought-together products
- Created interactive analytics dashboards with Recharts displaying real-time KPIs, trend charts, and category performance breakdowns
| Layer | Technology |
|---|---|
| Frontend | React 18, Vite, React Router, Recharts, Axios |
| Backend | Django 4.2, Django REST Framework, Simple JWT |
| Database | SQLite / MySQL |
| ML/AI | Scikit-learn, Pandas, NumPy |
| Auth | JWT (JSON Web Tokens) |