📊 FinTracker — AI-Powered Personal Finance Dashboard

FinTracker is a secure, full-stack web application designed to help users track, analyze, and predict their expenses.
It goes beyond traditional finance apps by integrating Machine Learning models and a Large Language Model (LLM) for personalized financial advice.

---

Deployed link

https://fin-tracker-hazel.vercel.app/

🚀 Key Highlights

• 📈 Predict future spending using ML (burn-rate forecasting)
• 🚨 Detect unusual expenses using anomaly detection
• 🤖 AI-powered financial assistant (FinBot)
• 🔐 Secure authentication with JWT
• 📊 Interactive dashboards and visualizations

---

🧠 Machine Learning Features

📈 Predictive Analytics (Burn Rate)

• Uses Linear Regression on daily spending data
• Forecasts end-of-month expenses
• Displays predicted trajectory on charts

🚨 Anomaly Detection

• Uses Isolation Forest
• Identifies unusual transactions (spikes/outliers)
• Helps detect unexpected expenses

🧹 Data Processing

• Outlier filtering using IQR (Interquartile Range)
• Clean time-based dataset for better predictions

---

🤖 FinBot — AI Financial Assistant

• Built using Google Gemini API (gemini-2.5-flash)
• Uses RAG (Retrieval-Augmented Generation)
• Injects real-time SQL insights into prompts

Provides:

• Spending insights
• Budget suggestions
• Smart recommendations

---

🏗️ Tech Stack

Frontend

• React.js
• React Router DOM
• Custom CSS (dark theme + animations)

Backend

• Python (Flask)
• Flask-JWT-Extended
• Flask-CORS

Database

• PostgreSQL
• psycopg2

Machine Learning

• pandas, numpy
• scikit-learn

Deployment

• Render

---

🔐 Security Features

• JWT-based authentication
• Password hashing (Werkzeug)
• Protected routes
• Input validation
• Chronological lock (no future transactions)

---

⚙️ Core Features

1. Secure Authentication

• User registration & login
• Encrypted passwords
• JWT-based sessions

2. Smart Transaction Management

• Add expenses with category & timestamp
• Real-time balance updates
• Strict validation system

3. Predictive Analytics

• Daily spending aggregation
• Monthly forecast
• Interactive charts

4. Anomaly Detection

• Flags unusual spending automatically

5. AI Assistant (FinBot)

• Context-aware financial advice
• Uses real-time database insights
• Fallback system for reliability

---

📊 System Architecture

User → React → Flask API → PostgreSQL
↓
ML Models (Regression + Isolation Forest)
↓
Gemini API (FinBot)

---

🚀 Future Improvements

• ARIMA / LSTM models
• Mobile app
• Bank API integration
• Budget planning features

---

🛠️ Setup

git clone <your-repo-url>

cd frontend
npm install
npm run dev

cd backend
pip install -r requirements.txt
python app.py

---