EquityLens

EquityLens is an advanced AI-powered continuous learning analytical platform for financial markets (Stocks, Commodities, and Cryptocurrencies). It leverages machine learning models (ARIMA, LSTM, CNN, Random Forest) to provide asset price predictions, paired with real-time market data directly integrated with Yahoo Finance. It also features explainable AI (XAI) using SHAP and LIME to interpret deep learning and ensemble predictions.

---

📸 Project Screenshots

Click to view screenshots!

(You can add more screenshots just by dropping them in the above section)

---

Key Features

• Multi-Asset Forecasting: Real-time and forward-looking predictions for Indian Equities (NSE), Gold, Silver, and Cryptocurrencies (BTC, ETH).
• Advanced ML Pipeline: Multiple prediction algorithms including Random Forest, ARIMA, LSTM, and simulated CNN/RNN.
• Explainable AI (XAI): Full transparency into what drives predictions using SHAP (feature importance) and LIME (local surrogate models).
• NIFTY 50 Clustering: Principal Component Analysis (PCA) and K-Means clustering for grouping large-cap Indian stocks by market behavior.
• Robust Market Data: Automatic multi-layer fallback strategy for yfinance to bypass cloud-provider IP rate limits.
• Continuous Background Caching: Prediction models pre-warm via cron schedules using Redis/LocMem, making frontend API hits lightning fast.

---

Tech Stack

Frontend:

• React (Vite)
• Tailwind CSS
• Recharts / Chart.js for visualization
• Axios for API integration

Backend:

• Python (Django 5.0+, Django REST Framework)
• Machine Learning: scikit-learn, tensorflow, statsmodels
• Market Data: yfinance, pandas, numpy
• Explainability: shap, lime

Infrastructure & Deployment:

• Gunicorn & Nginx
• Redis (Caching)
• PM2 (Process Manager)
• Azure Virtual Machines (Ubuntu)
• GitHub Actions (CI/CD Automated Deployment)

---

System Architecture & Sequence Diagram

The following sequence diagram illustrates the flow of data from the user requesting a prediction to the AI models serving cached or live generated data.

sequenceDiagram
    participant User as User (React Frontend)
    participant Nginx as Nginx (Reverse Proxy)
    participant API as Django REST API
    participant Cache as Redis/LocMem Cache
    participant YF as Yahoo Finance API
    participant ML as Machine Learning Pipeline
    
    User->>Nginx: GET /api/predict/?asset=RELIANCE.NS
    Nginx->>API: Forward Request
    
    API->>Cache: Check for pre-warmed prediction
    alt Cache Hit
        Cache-->>API: Return cached JSON prediction
        API-->>Nginx: HTTP 200 OK
        Nginx-->>User: Display Chart & Metrics
    else Cache Miss / Stale
        API->>YF: Fetch historical OHLC data
        alt Data Fetch Fails
            YF-->>API: rate-limited / error
            API->>YF: Fallback to Ticker.history() or ETF proxy
        end
        YF-->>API: Return historical DataFrame
        
        API->>ML: Pass DataFrame for Training & Prediction
        ML-->>API: Return Forecast Values + Metrics (RMSE, MAE)
        
        API->>Cache: Save new prediction (valid for 30m)
        API-->>Nginx: HTTP 200 OK (Fresh Data)
        Nginx-->>User: Display Chart & Metrics
    end

Loading

---

Local Setup and Installation

1. Clone the repository

git clone https://github.com/bhushantile20/EquityLens.git
cd EquityLens

2. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate  # Or `venv\Scripts\activate` on Windows
pip install -r requirements.txt

# Setup Database and Demo Data
python manage.py makemigrations
python manage.py migrate
python manage.py create_demo_user

# Run server
python manage.py runserver

3. Frontend Setup

cd frontend
npm install

# Set up local environment variables
echo "VITE_API_BASE_URL=http://localhost:8000/api/" > .env

# Run development server
npm run dev

---

Production Deployment workflow

This project utilizes purely automated CI/CD deployments through GitHub Actions perfectly suited for Azure.

When code is pushed to the main branch:

1. GitHub Actions checks out the repository.
2. Code is securely teleported to the Azure Server via SSH/SCP.
3. A remote script is triggered that:
   • Rebuilds the Vite React dashboard for production.
   • Updates Python dependencies and runs manage.py migrate.
   • Auto-restarts the PM2 Django process and Nginx reverse proxy.

For detailed local-to-server configurations, please ensure CORS, CSRF, and Allowed Hosts are populated in your .env correctly.