QuantifyAI

Overview

QuantifyAI is an ML-driven backtesting platform for equity trading strategies. It pulls live market data, computes classic technical indicators, trains a classifier to generate buy/sell signals, and runs portfolio simulations — all surfaced through an interactive Streamlit dashboard.

What was built

• Live data ingestion via yfinance — any ticker, any lookback window.
• Feature engineering pipeline computing RSI, MACD, and Bollinger Bands from raw OHLCV data.
• Random Forest classifier trained on technical features to produce buy/sell/hold signals; evaluated with ROC-AUC.
• Backtrader integration for portfolio simulation — applies model signals to historical data and tracks portfolio value over time.
• Streamlit dashboard for real-time analytics: model metrics, signal overlays, and backtest equity curves.
• requirements.txt and clean project layout for straightforward local setup.

Why it matters

Combining ML signal generation with a proper backtesting framework (rather than simple forward simulation) is a meaningful step toward realistic strategy evaluation. Backtrader handles slippage, position sizing, and trade execution mechanics that naive simulations ignore — making the results more honest about what the strategy would have actually returned.