Specializing in non-stationary time series modeling, probabilistic neural networks, and high-frequency statistical arbitrage. Deployed predictive models managing $450M in quantitative assets.
Verified model metrics evaluated on out-of-sample historical market data
Machine learning architectures and reproducible backtesting suites
Event-driven algorithmic backtesting framework in C++ and Python supporting non-linear slippage models and microstructure simulation.
Deep learning architecture for multi-horizon financial forecasting utilizing Gaussian process attention heads.
Available for quantitative fund advisory, predictive modeling audits, and machine learning architecture.