Projects
Libraries and engines for pricing, hedging, risk, and market simulation.
Machine Learning
Neural hedging policies trained against convex risk measures over simulated markets with transaction costs, accelerated by fused CUDA path kernels, plus a deep BSDE pricer.
Systems
A C++20 matching engine with strict price-time priority, a dense tick-ladder book with O(1) best-quote lookup, and allocation-free hot paths for sub-microsecond order processing.
Numerical Methods
An object-oriented engine pricing vanilla and exotic options across six stochastic processes, with analytic, lattice, finite-difference, Fourier, and Monte Carlo methods and a Streamlit dashboard.
Asset Allocation
A quant library for portfolio construction and risk, spanning hierarchical and clustered allocation, coherent and spectral risk measures, copula tail modelling, and maximum-likelihood SDE fitting.
Time Series
A controlled benchmark of Transformer, state-space, and temporal-convolution models against classical time-series baselines for short-horizon market forecasting, with matched training and ablations.
Generative Models
Generative models for synthetic asset-price paths, judged on market stylised facts and on recovering known stochastic-volatility dynamics, with a built-in benchmark from simulators whose moments are known.
