Stockify

AI-powered stock price prediction — Stockify uses LSTM neural networks to forecast next-day stock price movements. It predicts whether a stock will go UP, DOWN, or stay FLAT, along with the expected return percentage and confidence score. The project includes a FastAPI backend for model training and inference, and a Chrome browser extension for a seamless user experience.

Icon by srip

---

Features

• LSTM Ensemble Predictions — Trains 3 LSTM models with varying lookback windows (30, 60, and 90 days) and combines their outputs using confidence-weighted voting for more robust predictions.
• On-the-Fly Training — If no trained model exists for a requested stock symbol, the system automatically downloads historical data from Yahoo Finance and trains a model from scratch (~1 minute).
• Stock Search with Fuzzy Matching — Search for stocks by company name or ticker symbol. Uses rapidfuzz for prefix and fuzzy matching, returning the top 5 closest results.
• Chrome Extension UI — A Manifest V3 Chrome extension with a popup interface to search stocks, trigger predictions, and view results directly in your browser.

---

How It Works

1. Data Collection — Historical daily stock data (Open, High, Low, Close, Volume) is fetched from Yahoo Finance via the yfinance library.
2. Feature Engineering — Raw OHLCV data is combined with technical indicators to create a rich feature set for the model.
3. Sequence Generation — Data is converted into fixed-length sequences (30, 60, or 90 days) using a sliding window approach.
4. LSTM Training — A multi-layer LSTM network is trained on the sequences to predict the next day's price direction.
5. Ensemble Aggregation — Predictions from all 3 sequence lengths are combined using confidence-weighted voting.
6. Output — The system returns the predicted direction (UP/DOWN/FLAT), expected return percentage, predicted closing price, and model confidence.

---

Getting Started

Prerequisites

• Python 3.8+
• Google Chrome (for the browser extension)

Backend Setup

1. Clone the repository

   git clone https://github.com/Choco-10/Stockify.git
   cd Stockify

2. Install Python dependencies

   cd server
   pip install -r requirements.txt

3. Start the FastAPI server

   uvicorn app:app --reload

   The server will be available at http://127.0.0.1:8000.

4. Verify the server is running Open http://127.0.0.1:8000/docs in your browser to see the interactive API documentation (Swagger UI).

Chrome Extension Setup

1. Open Google Chrome and navigate to chrome://extensions/
2. Enable Developer mode (toggle in the top-right corner)
3. Click Load unpacked and select the stockify-extension/ folder
4. The Stockify icon will appear in your browser toolbar
5. Click the icon, search for a stock, and hit Predict

Note: By default the extension connects to the local server (http://127.0.0.1:8000). To use the production server, edit stockify-extension/config.js and uncomment the Render URL.

---

Disclaimer

Stockify is built for educational and research purposes only. It should not be used as financial advice. Stock markets are inherently unpredictable, and past performance does not guarantee future results. Always consult a qualified financial advisor before making investment decisions.