index.md

py_rt Trading System Documentation

Welcome to the comprehensive documentation for the py_rt (Python + Rust Trading) system - a high-performance algorithmic trading platform combining Python's ML ecosystem with Rust's speed and safety.

Quick Navigation

Getting Started

• Quick Start Guide - Get up and running in 10 minutes
• Installation - Detailed installation instructions
• Configuration Guide - System configuration
• First Strategy - Build your first trading strategy

Architecture

• System Overview - High-level architecture
• Component Design - Detailed component descriptions
• Data Flow - Message flow and communication
• Deployment - Production deployment guide

API Reference

• Python API - Python modules and classes
• Rust API - Rust crates and types
• REST API - REST endpoints
• WebSocket API - WebSocket protocol
• ZMQ Messages - ZeroMQ message formats

Development

• Contributing - How to contribute
• Code Style - Coding standards
• Testing - Testing strategies
• Performance Tuning - Optimization guide
• Troubleshooting - Common issues

User Guides

• Strategy Development - Build trading strategies
• Backtesting - Test strategies on historical data
• ML Integration - Integrate ML models
• Risk Management - Configure risk limits
• Monitoring - System monitoring and alerts

What is py_rt?

py_rt is a hybrid trading system that combines:

• Python for strategy development, ML models, and backtesting
• Rust for low-latency execution, risk management, and market data processing
• ZeroMQ for high-performance inter-process communication
• Alpaca Markets API for live trading and market data

Key Features

• Low Latency: Sub-5ms end-to-end order execution
• Type Safety: Compile-time guarantees across Rust and Python
• ML Ready: PyO3 bindings for seamless Python ML integration
• Risk Controls: Pre-trade risk checks and circuit breakers
• Observability: Prometheus metrics and structured logging
• Scalable: Microservices architecture scales horizontally

System Components

graph TD
    A[Market Data Service] -->|ZMQ| B[Signal Bridge]
    B -->|ZMQ| C[Risk Manager]
    C -->|ZMQ| D[Execution Engine]
    D -->|REST| E[Alpaca Markets]
    E -->|WebSocket| A

Loading

1. Market Data Service (Rust) - Real-time market data ingestion
2. Signal Bridge (Rust + Python) - ML-based signal generation
3. Risk Manager (Rust) - Pre-trade risk checks and position tracking
4. Execution Engine (Rust) - Smart order routing and execution

Performance

• Market Data Processing: 50,000+ updates/second
• Order Book Updates: <50μs latency (p99)
• Signal to Execution: <5ms end-to-end
• Memory Footprint: <100MB per component

Technology Stack

Python Stack

• ML/Data: pandas, numpy, scipy, scikit-learn
• Backtesting: Custom engine with walk-forward analysis
• API: alpaca-py, pydantic
• Testing: pytest, hypothesis

Rust Stack

• Runtime: Tokio async
• Messaging: ZeroMQ
• Serialization: serde
• API: reqwest, tungstenite
• Testing: cargo test, proptest

Documentation Structure

This documentation is organized into five main sections:

1. Guides - Step-by-step tutorials for common tasks
2. Architecture - System design and technical decisions
3. API Reference - Complete API documentation
4. Developer - Contributing and development guides
5. Examples - Code examples and use cases

Support and Community

• GitHub Issues: Report bugs and request features
• GitHub Discussions: Ask questions and share ideas
• Documentation: You're here!

License

Licensed under the Apache License 2.0. See LICENSE for details.

Author

Davi Castro Samora

• GitHub: @SamoraDC

---

Version: 0.1.0 | Last Updated: 2025-10-14