QUANT
ARTISAN
Algorithmic trading strategies built with the precision of a craftsman and the autonomy of an AI agent.
Autonomous Strategies. Handcrafted Edge.
WHERE QUANTITATIVE
RIGOR MEETS
AGENTIC AUTONOMY
I build algorithmic trading systems that operate at the intersection of quantitative finance and autonomous AI. Each strategy is hand-crafted from first principles — not assembled from off-the-shelf components — and deployed through purpose-built agentic frameworks.
The artisan philosophy is deliberate: in a world of commoditized signals and crowded factors, edge comes from depth of craft — from understanding the microstructure, the regime, and the agent's decision boundary in ways that automated pipelines cannot replicate.
THE FOUR
FOUNDATIONS
Every system I build rests on these four pillars. Remove any one and the architecture collapses.
Agentic Intelligence
Autonomous AI agents that perceive market microstructure, reason over multi-timeframe signals, and execute with sub-second precision — without human intervention.
Quantitative Craft
Every strategy is hand-built from first principles — statistical edge identification, rigorous walk-forward validation, and robust out-of-sample testing before a single dollar is risked.
Systematic Risk
Position sizing, correlation management, and drawdown controls are not afterthoughts — they are the architecture. Capital preservation is the first rule of compounding.
Continuous Learning
Markets evolve. Strategies decay. Agents that adapt — through online learning, regime detection, and parameter recalibration — maintain edge where static systems fail.
FROM SIGNAL
TO EXECUTION
Signal Research
Identify statistically significant market anomalies across equities, crypto, and derivatives using factor analysis and ML feature importance.
Agent Design
Architect autonomous agents with defined perception, reasoning, and action loops. Each agent is a self-contained decision system.
Backtesting & Validation
Rigorous walk-forward testing, Monte Carlo simulation, and regime-conditional analysis to distinguish genuine edge from data-mining bias.
Live Deployment
Paper trading → live deployment with fractional sizing. Real-time monitoring, automated circuit breakers, and performance attribution.
LIVE SIGNALS &
RESEARCH
A selection of strategies in production and active research. Performance figures are audited trailing 12-month results.
Momentum Regime Agent
A multi-timeframe momentum strategy with an LLM-powered regime classifier. The agent switches between trend-following and mean-reversion modes based on macro context.
Microstructure Scalper
Order-book imbalance detection with sub-100ms execution. The agent reads Level 2 data, identifies short-term price pressure, and exits within seconds.
Volatility Surface Arb
Statistical arbitrage across the implied volatility surface. Identifies mispricings in term structure and skew, hedging delta and gamma continuously.
LET'S BUILD
EDGE TOGETHER
Whether you're looking to collaborate on a strategy, discuss agentic AI in finance, or simply exchange ideas — I'm open to conversations with serious practitioners.