Building resilient software and secure systems.

At NYU’s OSIRIS Lab, I study indirect prompt injection in AI agents. A document or tool result can contain instructions an agent should not follow. My research looks at when to add security checks and how those checks affect task completion, latency, and computational cost.

Quantitative development interests me because it brings software engineering together with probability and statistical reasoning. I want to understand how those methods are used to build financial software. In my blackjack simulator, I’ve explored expected returns and how sampling uncertainty affects the conclusions you can draw from a simulation.

On campus, I’m involved in Tech@NYU, Chess Team, Robotics Club, Stern Finance Society, and Philosophy Club.