StockVision Pro

StockVision Pro is an AI-powered stock prediction system that analyzes historical market data and provides next-day price predictions with technical analysis for multiple stock tickers.

---

Features

• Multi-stock prediction — Supports AAPL, AMZN, GOOGL, META, MSFT, TSLA
• LSTM-based deep learning models — Individual models trained per ticker
• Automatic technical indicators — Calculates MA20, MA50, MA100, RSI
• Real-time market data — Fetches live data via Yahoo Finance API
• Next-day price prediction — AI-powered forecasting
• BUY / SELL / HOLD recommendations — Automated trading signals
• Interactive charts — Visualizes actual vs predicted prices
• REST API + Web dashboard — Full-stack Flask application

---

Project Structure

StockVision_Pro/
│
├── app.py                 # Main Flask backend
├── api/                   # Vercel serverless entry
│   └── index.py
├── models/                # Trained LSTM models (.keras)
├── scalers/               # Feature scalers (.pkl)
├── templates/             # HTML frontend
│   └── index.html
├── requirements.txt
├── vercel.json
└── .gitignore

---

Tech Stack

Layer	Technology
Language	Python
Web Framework	Flask
Deep Learning	TensorFlow / Keras
Data Processing	NumPy & Pandas
Feature Scaling	Scikit-learn
Market Data	Yahoo Finance (yfinance)
Deployment	Render

---

Running Locally

1. Install dependencies:

pip install -r requirements.txt

2. Start the server:

python app.py

3. Open in browser: http://localhost:5000

---

Output

Each prediction response includes:

Field	Description
Current Price	Latest market price
Predicted Price	Next-day AI forecast
Price Change	Absolute & percentage change
Recommendation	BUY / SELL / HOLD signal
Technical Indicators	MA20, MA50, MA100, RSI
Chart Data	Historical + predicted prices

---

Deployment

StockVision Pro is deployed on Render with continuous deployment from GitHub. The Flask API runs as a web service, and both frontend and backend are managed from a single repository.

---

Notes

• Models and scalers are loaded dynamically per ticker
• Files are cached in memory for faster performance
• Model files are stored in /models
• Scalers are stored in /scalers

---

License

This project is for educational and research purposes only. It does not provide financial or investment advice.

---

Authors

Arsalan Tahir · Usayd Arsalan · Bazyl Sheikh

---

Disclaimer: This tool is for educational purposes only. Always do your own research before making investment decisions.