Business Intelligence Visualization Tool for Algorithmic Trading

A comprehensive, open-source business intelligence visualization tool designed for algorithmic trading systems. This application transforms complex trading data into intuitive visualizations, enabling traders and analysts to make data-driven decisions.

Features

• Multi-Asset Class Analysis: Support for equities, cryptocurrencies, and forex
• Interactive Dashboards: Dynamic charts and filters for data exploration
• Strategy Performance Tracking: Monitor and compare trading strategies
• Multi-Dimensional Analysis: Filter by symbol, strategy, time period, and asset class
• Real Market Data Integration: Uses real-world data from Yahoo Finance
• Star Schema Data Model: Optimized for analytical processing

Technology Stack

• Backend: FastAPI + Uvicorn
• Frontend: Streamlit + Plotly
• Database: SQLite with SQLAlchemy ORM
• Data Processing: Pandas + yfinance
• Version Control: Git

Getting Started

Prerequisites

• Python 3.8+
• pip
• virtualenv (recommended)

Installation

1. Clone the repository

   git clone https://github.com/your-username/trading-bi-tool.git
   cd trading-bi-tool

2. Create and activate a virtual environment

   python -m venv ProjectVenv
   # On Windows
   ProjectVenv\Scripts\activate
   # On macOS/Linux
   source ProjectVenv/bin/activate

3. Install dependencies

   pip install -r requirements.txt

Running the Application

1. Start the FastAPI backend server

   python bi_tool_test.py

2. In a new terminal, start the Streamlit frontend

   streamlit run bi_tool_streamlit.py

3. Load sample data (if needed)

   python load_market_data.py

4. Access the dashboard at http://localhost:8501

Project Structure

Project/
├── Backend/
│   ├── __init__.py
│   ├── database.py      # Database models and connection
│   ├── etl.py           # ETL processes
│   ├── main.py          # API entry point
│   ├── models.py        # Data models
│   ├── routes.py        # API routes
│   └── services.py      # Business logic
│
├── Frontend/
│   ├── __init__.py
│   ├── app.py           # Streamlit application
│   ├── charts.py        # Chart definitions
│   └── data_loader.py   # API integration
│
├── ProjectVenv/         # Virtual environment
├── bi_tool_streamlit.py # Streamlit dashboard
├── bi_tool_test.py      # FastAPI server
├── database.py          # Main database models
├── load_market_data.py  # Data loading script
├── data.db              # SQLite database
└── README.md            # Project documentation

Core Components

1. Data Model

The system implements a star schema with:

• Fact Table: trade_facts
• Dimension Tables: symbol_dim, trader_dim, strategy_dim

This design enables efficient querying and analysis across multiple dimensions.

2. API Endpoints

• /symbols - List all trading symbols
• /strategies - List all trading strategies
• /traders - List all traders
• /trades - Get trade data with optional filtering
• /performance/{strategy_id} - Get performance data for a strategy

3. Dashboard Visualizations

• Trading value by symbol
• Trading distribution by strategy
• Performance over time
• Recent trades table
• Key performance metrics

Contributors

• Diego Mckay - Initial project setup, database foundation, API framework, basic dashboard
• Shefreen Kaur - Advanced data modeling, data pipeline, enhanced API layer, advanced visualizations, system integration

Academic Context

This project was developed as part of the COMP 430 course. It demonstrates the application of business intelligence principles, data visualization techniques, and full-stack development in the context of algorithmic trading systems.

Future Enhancements

• Real-time data streaming via WebSockets
• User authentication and multitenancy
• Strategy backtesting capabilities
• Machine learning integration for predictive analytics
• Export capabilities for reports and analysis