Projects & Work

Things I've built, researched, or contributed to.

Work

Lead ML Engineer — Stealth Startup

Mar 2025 – Present

Building the signal processing and ML pipeline for a novel passive RF sensing system deployed on the frontline in Ukraine.

Signal ProcessingMLRF SensingDefense

Built the experimentation and causal inference evaluation framework used on the world's largest online donation platform (10M+ daily users) into a highly flexible tool allowing data scientists to answer complex questions about user behavior.

Causal InferenceExperimentationMLPython

Trained a generalized fraud detection model and built the pipeline that — with minor modifications — to this day processes all healthcare claims at three of the top five largest health insurers in Germany. This became QuantCo's first product suite (Themis, then Aramis).

Fraud DetectionMLHealthcarePython

Projects

Exploratory theoretical work on air defense. Develops a process model of interceptor group response time and a game-theoretic model of high-level AD placement — a defender with limited resources against an attacker that learns over time.

DefenseOperations ResearchGame TheoryPDF ↗

Python toolkit for simulating Time Difference of Arrival (TDOA) source localization. Given a sensor array, recovers an emitter's position from differences in signal arrival times. Implements closed-form and Levenberg-Marquardt solvers, Monte Carlo error analysis, and GDOP spatial mapping via the Fisher Information Matrix.

Signal ProcessingLocalizationPythonRFGitHub

Simulation framework for pricing a widget on an online marketplace. Implements and evaluates random baseline, epsilon-greedy multi-armed bandit, and UCB1 against arbitrary demand functions.

BanditsPricingPythonGitHub

Research

Designed and analyzed an economic mechanism to disincentivize misinformation. Class project for CS136 (Economics and Computation), advised by Prof. David Parkes. Co-authored with Tejovan Parker and Jacob Cremers.

Mechanism DesignEconomicsMisinformationPDF ↗