University of Pennsylvania. I specialize in building full-stack systems, large language models, and the occasional passion project that keeps me up at night.
Tech has been part of my life for as long as I can remember. As a toddler, I was dismantling remotes and rewiring toys. By middle school, I was building computers and teaching myself to code. At 18, online poker led me to discover a vulnerability in a crypto casino's shuffling algorithm that created predictable patterns every few hundred hands. I taught myself to code and developed a program to exploit it (ethically, of course), and was hired to help fix their system. Since then, I've been obsessed with finding ways to apply algorithmic reasoning—through machine learning, game theory, and probabilistic thinking—to real-world conflict and decision-making.
I'm excited by projects that are ambitious, creative, and technically challenging. I enjoy turning complex ideas into intuitive and functional digital experiences, whether it's a rough concept or a fully formed vision. When I'm not studying or coding, you'll find me at the gym finishing training with my MMA team or preparing for Marines Officer Candidates School next summer. I also like to play poker (my main source of income), and eat hotpots (my main source of expenses).
I'm currently working on projects in artificial intelligence and blockchain/Web3 technologies. My research interests include machine learning applications in tech, algorithmic trading strategies, and the mathematical foundations of cryptographic systems.
In the past, I've always found great success in collaborating and connecting with new people. I'd love to connect, so don't hesitate to reach out.
Python
C++
R
OCaml
Java
MATLAB
HTML5
CSS
JavaScript
NumPy
Pandas
Scikit-learn
PyTorch
React
TensorFlow
Matplotlib
MySQL
PostgreSQL
MongoDB
Node.js
Git
Built Python ETL pipelines to automate processing of 50,000+ financial records and developed regression models achieving 92% accuracy in tax liability predictions.
Built NLP text classification system to categorize 1,000+ legal documents, reducing manual review time by 50%, and conducted statistical analysis on settlement data to inform case strategy.
Supported major financial transactions and market analysis for billion-dollar deals while conducting regulatory impact assessments and client presentations.
Working under Professor Michael C. Horowitz to digitize and analyze over 50 years of historical U.S. military appropriations data, developing Python-based OCR and data extraction pipelines.
Built Monte Carlo simulations and predictive statistical models for tokamak fusion reactors, automating data pipelines that reduced analysis time by 75%.