Alpha Hunter — Autonomous AI Trading Scientist

Alpha Hunter is an autonomous AI trading scientist built to close the loop between quantitative research, strategy validation, portfolio allocation, and execution. Instead of simply asking an AI what to buy, Alpha Hunter first asks whether a trading strategy actually has a robust edge. The system discovers strategy hypotheses and evaluates them using market data, historical performance, out-of-sample validation, robustness analysis, and market-regime compatibility. An adversarial AI layer then attempts to break promising strategies by looking for weaknesses such as overfitting, instability, regime dependence, concentration, and poor generalization. Surviving strategies receive an Edge Score and can compete for capital through Alpha Hunter's adaptive allocation engine. Allocation considers strategy quality, risk, volatility, portfolio exposure, correlation, market regime, and current opportunity rather than blindly increasing every position. Alpha Hunter supports equities, ETFs, and crypto. Crypto is not permanently excluded; when its risk-adjusted opportunity deteriorates, the system can pause new allocation while continuing to monitor existing positions. A deterministic risk engine acts as a safety layer between AI recommendations and execution. It enforces portfolio and position constraints while allowing profitable positions to use adaptive trailing protection designed to protect accumulated gains and let strong trades continue running. The execution layer connects to Alpaca Paper Trading. Current positions, quantities, orders, P&L, and broker execution states are reconciled with Alpaca rather than being simulated in the UI. The result is a complete autonomous trading research loop: Discover → Test → Challenge → Score → Allocate → Execute → Monitor → Learn. Alpha Hunter is designed not merely to generate trades, but to continuously determine which strategies deserve trust and where capital has the best risk-adjusted opportunity.