QuantPilot-AI

QuantPilot-AI is a beginner-friendly stock research and backtesting platform.

The goal of this project is to help users learn how stock data, technical indicators, trading signals, backtesting, performance metrics, and simple research reports can work together in one explainable workflow.

This project does not try to predict stock prices with certainty. It is built for education, experimentation, and research.

Current V1 Features

• Sample stock K-line data from CSV
• Technical indicators:
  • MA5
  • MA20
  • RSI
  • CCI
• MA crossover signal generation
• Simple long-only backtesting
• Performance metrics
• Rule-based strategy report
• One-command workflow using:

python src/main.py

Project Structure

QuantPilot-AI/
├── data/
│   └── sample/
│       ├── README.md
│       └── sample_stock.csv
├── docs/
│   ├── project-plan.md
│   └── stock-research-template.md
├── src/
│   ├── backtester.py
│   ├── data_loader.py
│   ├── indicators.py
│   ├── main.py
│   ├── metrics.py
│   ├── report_generator.py
│   └── strategy.py
├── .gitignore
└── README.md

Installation

Clone the project and enter the project folder:

git clone <your-repository-url>
cd QuantPilot-AI

Install the required Python package:

pip install pandas

Python 3.10 or newer is recommended.

Usage

Run the full V1 workflow from the project root:

python src/main.py

This command will:

1. Load data/sample/sample_stock.csv
2. Add MA5, MA20, RSI, and CCI indicators
3. Generate MA crossover signals
4. Run a simple long-only backtest
5. Calculate performance metrics
6. Print a rule-based strategy report
7. Print the last 10 rows of the backtest result

Example Output

The current sample data produces output similar to:

initial_value: 10000.0
final_value: 11227.91
total_return_pct: 12.28
max_drawdown_pct: -5.78
buy_signals: 1
currently_holding: True

The exact formatting may differ slightly, but the V1 result should show one buy signal and a final portfolio value around 11227.91.

Disclaimer

This project is for educational and research purposes only. It is not financial advice.

V2 Roadmap

• Real stock data loader
• Streamlit dashboard
• Better risk metrics
• AI-assisted report generation
• Multi-agent research system