Research

A record of my research and published work, spanning my academic years at Princeton and my time leading AI teams in industry.

My Ph.D. research focused on AI for Security and Security of AI. I also contributed to research on anonymity at the network level, adversarial examples in computer vision, and secure communication channels for constrained devices.

I later spent years at Google as an AI Lead across Search, Recommendations, Agents, and Enterprise Agents, shipping AI systems at scale to real users. That experience shaped my research and engineering taste. I have since left Google and I’m now building a company focused on AI Security.

Active academic research output has paused, but the history stays here.

Recently, my work has been broadly focused on the security of multi-step AI agents and multi-agent environments, with a core emphasis on adversarial detection, alignment, and remediation, and the development of defensive primitives for AI agents.

For an always-current list of publications, see my Google Scholar profile.