Skip to content

architecture.md

Latest commit

 

History

History
96 lines (75 loc) · 3.5 KB

architecture.md

File metadata and controls

96 lines (75 loc) · 3.5 KB

🏗️ Architecture

TraderMagic is built with a modern, service-based architecture that separates concerns and provides flexibility.

🧩 Components

The system consists of four main services that work together:

📡 Data Retrieval Service

Purpose: Fetches technical indicators from external APIs
Key Features:

  • Polls TAAPI.io for RSI (Relative Strength Index) data
  • Rate-limiting to respect API constraints
  • Caches results in Redis for other services

Technology:

  • Python async HTTP client
  • TAAPI.io API integration
  • Configurable polling intervals

🧠 AI Decision Engine

Purpose: Analyzes market data and makes trading decisions
Key Features:

  • Uses locally-run Ollama LLM to analyze RSI data
  • Determines whether to buy, sell, or hold based on analysis
  • Explains reasoning behind each decision

Technology:

  • Ollama integration for local LLM inference
  • Prompt engineering for financial analysis
  • Redis for receiving data and publishing decisions

💹 Trade Execution Service

Purpose: Executes trades based on AI decisions
Key Features:

  • Connects to Alpaca for trade execution
  • Implements trade amount calculation
  • Handles order placement and tracking
  • Respects trading enabled/disabled flag

Technology:

  • Alpaca API integration
  • Portfolio management logic
  • Trading safeguards and limits

🖥️ Web Dashboard

Purpose: User interface for monitoring and control
Key Features:

  • Real-time trade monitoring
  • Trading control interface
  • Trade mode configuration
  • Theme customization

Technology:

  • Flask for backend API
  • Socket.IO for real-time updates
  • Responsive frontend design

🔄 Data Flow

  1. Data Retrieval Service polls TAAPI.io for RSI data
  2. Data is stored in Redis with symbol-specific keys
  3. AI Decision Engine consumes RSI data and generates trade signals
  4. Signals are stored in Redis with corresponding keys
  5. Trade Execution Service processes signals and executes trades
  6. Trade results are stored in Redis
  7. Web Dashboard displays all data from Redis in real-time

📦 Containerization

All components are containerized using Docker:

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│ Data Retrieval  │    │  AI Decision    │    │Trade Execution  │
│    Service      │    │    Engine       │    │    Service      │
└────────┬────────┘    └────────┬────────┘    └────────┬────────┘
         │                      │                      │
         │                      ▼                      │
         │             ┌─────────────────┐             │
         └────────────►│      Redis      │◄────────────┘
                       │                 │
                       └────────┬────────┘
                                │
                                ▼
                       ┌─────────────────┐
                       │  Web Dashboard  │
                       │                 │
                       └─────────────────┘

Each service can be scaled independently and communicates via Redis.