What I'm Working On

Current Projects

Brawl Stars Reinforcement Learning An RL agent and computer-vision pipeline that plays Brawl Stars at a competitive level, trained with PPO across ~450M simulated matches.
Senior Thesis: How Hackathons Prepare Students for Tech Careers A two-year qualitative study of how hackathon participation feeds into startup formation and the tech industry.

Brawl Stars Reinforcement Learning

Jul. 2026 – Present Personal project Python · PyTorch · RL

I'm building a reinforcement learning agent that plays the mobile game Brawl Stars at a competitive level, paired with a computer-vision pipeline that reads game state straight from raw frames.

Wrote a custom vectorized simulator that mirrors Brawl Stars' mechanics so I can run huge batches of matches in parallel instead of being bottlenecked by the real game.
Training the policy with Proximal Policy Optimization (PPO) — roughly 450 million simulated matches per training run.
Documenting each stage of development in a YouTube series, which has passed 500,000 views.

Full write-up GitHub YouTube series

↑ Back to top

Senior Thesis: How Hackathons Prepare Students for Tech Careers

SOCL217 & SOCL400 Sep. 2025 – May 2027 Qualitative research

A two-year-long qualitative research project examining how hackathon participation relates to startup development and the broader tech industry.

Participatory observation of 7+ hackathons across the East Coast.
Semi-structured interviews with hackathon organizers and sponsors.
Preparing findings toward the final thesis due May 2027.

If you're an organizer, company, or VC fund that sponsors hackathons, I'd love to hear from you about your experience and insights — feel free to reach out.

↑ Back to top