Autonomous AI Edge-Finding Agent
An autonomous AI agent that finds statistical edges in prediction markets, manages risk with deterministic Kelly, and learns from settled trades
Openthomas is an autonomous AI trading agent designed for prediction markets including Polymarket and Kalshi. It continuously scans markets to find statistical edges using various data sources and machine learning models. The agent is fully autonomous and paper-trades by default to validate its strategies.
Entry: The agent identifies markets where it has a demonstrable statistical edge based on historical patterns, real-world data feeds, or modeling. It sizes positions using deterministic Kelly Criterion, ensuring consistent risk management. The system learns from every settled trade, updating its models and edge calculations.
Exit: Positions are held until settlement, with the agent capable of hedging or closing early if new information changes the expected value. The implementation includes a frontend for monitoring and manual override, but the core decisions are fully automated.