Market Forecasting
LSTM and Transformer models on limit order book features.
Research · Engineering · Quantitative Finance
Research and applied work in quantitative finance.
I build quantitative models and the software systems that make them useful — from market microstructure simulations to production financial infrastructure.
Seeking Fall 2026 internships
Section 01
Publications & articles
Hinges and Thresholds: How a ReLU Thinks Like a Call Option
A ReLU in a neural network and a call option at expiration look like they belong to different worlds. One is taught in machine learning; the other is traded in financial markets. Yet their core rule is identical: below a threshold, the payoff is zero. Above it, the response increases one-for-one.
Learning the Whole Volatility Surface
Most machine-learning option-pricing models treat contracts as independent rows. Markets do not. A strike belongs to a smile, a maturity belongs to a term structure, and every prediction must coexist with the rest of the surface without creating arbitrage.
Simulating the Glosten-Milgrom Model in the Browser
A walkthrough of the Glosten-Milgrom sequential trade model and its implementation.
Impact of US and EU Sanctions on the Russian Economy
Applied econometric techniques to analyze the macroeconomic consequences of US and EU sanctions on Russia. Research paper for my BSc thesis at Bocconi University.
Section 02
Research projects
Quant Sim
Interactive financial models in Rust and WebAssembly.
Portfolio Optimization Suite
Mean-variance and Black-Litterman implementations from scratch.