AI & Automation

Prediction Market Trading Bot

PythonNext.jsReactSupabaseRailwayVercelREST APIsRecharts

What It Does

A Python-based automated trading system for prediction markets. The bot runs multiple strategies that pull real-time data from external sources (NOAA weather forecasts, ESPN live sports probabilities) and compare them against prediction market prices. When the bot detects a gap between what the data says and what the market is pricing, it places trades automatically. It supports both paper trading and live execution, with a full risk management layer handling position sizing, daily loss limits, and cooldowns. All state is persisted to Supabase so the system survives restarts. A separate Next.js dashboard polls the database and displays open positions, P&L, trade history, and strategy performance in real time.

Key Features

  • Multiple automated strategies comparing external data sources against market prices
  • Weather strategy: NOAA forecasts vs. market-priced temperature and precipitation contracts
  • Sports strategies: ESPN live and pre-game win probabilities vs. market prices
  • Full risk management: position sizing, daily loss limits, cooldowns, and longshot protection
  • Paper trading and live execution modes
  • Real-time dashboard showing open positions, P&L, trade history, and strategy performance
  • All state persisted to Supabase for crash recovery and restart resilience
  • Remote kill switch via dashboard

Why I Built It

Prediction markets are a data problem. If you can get better data faster than the market is pricing, there are opportunities. I wanted to build a system that finds and acts on those gaps automatically.

What I Learned

Learned how to build a stateful trading system with crash recovery, integrate multiple real-time data APIs, implement risk management controls, and design a live monitoring dashboard for an autonomous system.

Skills Used & Gained

Trading SystemsAPI IntegrationRisk ManagementReal-Time Data ProcessingFull-Stack DevelopmentDatabase Design