💸 Smart Expense Analyzer & Fraud Risk Detector

An AI-powered full-stack financial analytics system that classifies expenses, detects fraudulent transactions, and forecasts future spending.

---

🚀 Features

• 🔍 Expense Classification – Categorizes transactions using a trained ML model
• 🚨 Fraud Detection – Detects anomalies using Isolation Forest
• 📈 Spending Forecast – Predicts future expenses using LSTM
• 📊 Dashboard Analytics – Visual insights of user spending
• 🔐 Authentication System – Secure login/signup with JWT

---

🛠️ Tech Stack

Frontend: React (Vite) Backend: FastAPI Database: PostgreSQL (Render) ML Models: Scikit-learn, PyTorch Deployment:

• Backend → Render
• Frontend → Vercel

---

🌐 Live Demo

• 🔗 Frontend: https://smart-expense-analyzer-and-fraud-ri.vercel.app/
• 🔗 Backend API: https://expense-backend-c63k.onrender.com/docs

---

📂 Project Structure

project-root/
│
├── backend/          # FastAPI backend
├── frontend/         # React frontend
├── models/           # ML models (.pkl, .pt)
├── requirements.txt
└── README.md

---

⚙️ Setup Instructions

1️⃣ Clone repo

git clone https://github.com/your-username/your-repo.git
cd your-repo

---

2️⃣ Backend setup

cd backend
pip install -r requirements.txt
uvicorn main:app --reload

---

3️⃣ Frontend setup

cd frontend
npm install
npm run dev

---

4️⃣ Environment Variables

Create .env file:

DATABASE_URL=your_postgres_url
SECRET_KEY=your_secret_key

---

🧠 ML Models

• Classifier: Expense category prediction
• Fraud Model: Isolation Forest for anomaly detection
• LSTM Model: Time-series forecasting

---

📌 Future Improvements

• Use real-world financial datasets
• Improve model accuracy
• Add explainability (why fraud detected)
• Enhance UI/UX

⭐ If you like this project, give it a star!