Hacking Capital is a production-ready, deterministic trading agent platform that combines advanced technical analysis with AI-powered pattern recognition for autonomous financial decision-making. The multi-agent system employs ensemble decision-making from specialized trading agents (trend-following, momentum, volatility, volume) that dynamically weight signals through configurable parameters, enabling sophisticated quantitative trading strategies without relying on large language models.
Live Demo Status: The application features a fully functional real-time paper trading environment with Server-Sent Events streaming, allowing users to interactively adjust agent weights and observe live market simulations. The demo is deployable and stable, running on containerized infrastructure with automated health monitoring and comprehensive error handling.
Core Finance Flow: Market data flows from Alpha Vantage API through technical indicator analysis (SMA, RSI, MACD, Bollinger Bands, etc.) and vector-based historical similarity matching, generating autonomous trade decisions with realistic transaction costs (0.1%), position sizing, and portfolio management. The system executes trades deterministically based on quantitative signals and ensemble weights.
Autonomy Level: Trading decisions are completely automated with real-time adjustable agent weightings accessible through the Streamlit interface. The system operates independently during trading hours, making deterministic decisions based on technical indicators and historical pattern matching without requiring human intervention.
Measurable Outcomes: Comprehensive performance analytics including Sharpe ratio, maximum drawdown, total return percentage, buy-and-hold comparisons, and detailed trade logs with transaction cost analysis. Users can quantify strategy effectiveness across different market conditions and time periods.
Judging Criteria Satisfaction
Running Code (Deployable Demo & Stability)
The application is fully containerized with Docker and includes automated health checks, comprehensive testing (unit and integration), and database persistence. The demo supports real-time paper trading with Server-Sent Events streaming, enabling judges to interact with live trading simulations. Production deployment scripts support multiple platforms including LiquidMetal AI Raindrop PaaS, Docker, Render, and Railway, with environment variable management and error handling for stability.
Innovation & Creativity (Novel Algorithms & Integrations)
The system introduces several novel approaches: vector-based historical similarity search using L2 distance matching in SQLite for pattern recognition, dynamic agent ensemble weighting with real-time configurability, and LLM-free deterministic trading that achieves sophisticated decision-making through quantitative signals. The architecture enables pluggable agent systems with specialized trading strategies that can be combined and weighted dynamically.
Real-world Impact (ROI & Problem Solved)
Addresses the critical challenge of algorithmic trading accessibility by providing a transparent, explainable quantitative trading platform. The system enables users to achieve measurable ROI improvements through data-driven decision-making rather than emotional trading, with comprehensive risk management metrics and performance benchmarking against buy-and-hold strategies. It democratizes access to professional-grade quantitative analysis tools.
Theme Alignment (Multi-asset Reasoning, APIs, Execution)
Multi-asset Reasoning: Designed to handle any asset with OHLCV data through flexible API integration
APIs: Seamlessly integrates with Alpha Vantage API for real-time and historical market data feeds
Execution: Implements complete trade execution simulation with realistic transaction costs, position management, and portfolio tracking
Technologies, Frameworks, Libraries, APIs, and Tools Used
Core Technologies
Python 3.11+: Primary programming language with async capabilities
FastAPI: High-performance REST API framework for backend services
Streamlit: Interactive web UI for real-time trading dashboard
Uvicorn: ASGI server for FastAPI deployment
Data Processing & Analysis
NumPy: Efficient numerical computations for technical indicators
Pandas: Data manipulation and time series analysis
SQLAlchemy: ORM for database operations
SQLite: Lightweight database with vector search capabilities
Infrastructure & Deployment
Docker & Docker Compose: Containerization and multi-service orchestration
Redis 7-alpine: High-performance caching layer
LiquidMetal AI Raindrop CLI: Primary PaaS deployment platform
Render: Alternative cloud deployment platform
Railway: Alternative cloud deployment platform
APIs & External Services
Alpha Vantage API: Real-time and historical market data provider
OpenAI API: Optional development tooling
Development & Quality Tools
Pytest: Comprehensive test suite with async support
Ruff: Fast Python linter and code formatter
MyPy: Static type checking
Homebrew: macOS package manager
Git: Version control system
Makefile: Build automation
Additional Libraries
Pydantic: Data validation and settings management
HTTPX: HTTP client library
Python-dotenv: Environment variable management
SSE-Starlette: Server-Sent Events for real-time streaming
Live Endpoints & Deployment Notes
The application is production-ready and can be deployed immediately using multiple methods:
Local Development:
UI

Quick start with Dockerdocker compose up –build -d# Or manual setupuvicorn app.main:app –reload –host 0.0.0.0 –port 8000 # APIstreamlit run ui/App.py –server.port 8501 –server.address 0.0.0.0 # UI

Live Endpoints:
Trading Dashboard UI: http://localhost:8501
API Documentation: http://localhost:8000/docs
Health Check: http://localhost:8000/api/v1/health
Cloud Deployment:
Primary: ./deploy.sh raindrop (LiquidMetal AI Raindrop PaaS)
Alternatives: ./deploy.sh docker, ./deploy.sh render, ./deploy.sh railway
Required Environment Variables:
ALPHAVANTAGE_API_KEY: Market data access
OPENAI_API_KEY: Optional development tooling
DATABASE_URL: SQLite connection string
REDIS_URL: Caching layer configuration
The system demonstrates enterprise-grade architecture with proper separation of concerns, comprehensive error handling, and scalable design patterns suitable for real-world quantitative trading applications.
This description is approximately 950 words and provides comprehensive coverage of all judging criteria with specific examples from your codebase implementation. It explicitly addresses each criterion while showcasing the technical sophistication and market readiness of your trading platform.