I’m a research fellow at HAI and the Stanford Intelligent Systems Laboratory (SISL) funded by the Schmidt Sciences Foundation. Prior to this I built fun things at Amazon and Capella Space. I completed my PhD in Aerospace Engineering at Stanford under Mykel Kochenderfer. My work focuses on building robust decision-making into autonomous robotic systems. More than anything, I love solving hard problems with good people.
My current research focuses on world-modeling, automated red-teaming, and autonomous decision making for space systems. I have active research projects in cybersecurity, robotics, and spacecraft operations. I currently have a strong interest in developing world-model based-controllers to enable robots to learn how to accomplish complex tasks from a small number of experiences. My work generally combines a mix of techniques from optimization, decision-making, reinforcement learning, and control theory with validation technicques and real-world deployments to realize robust autonomy to push exploration frontiers.
I was a founding engineer at Capella Space, the first US commercial synthetic aperture radar satellite imaging constellation. There I led the spacecraft operations group that successfully developed a fully-automated constellation operations system capable of delivering global high-resolution imagery on-demand without any human involvement. I subsequently led the constellation operations and space safety group at Project Leo (Kuiper). Most recently, I was a Principal Applied Scientist at Amazon Web Services, where I worked on building software services for large-scale distributed edge compute applications.
Astrodynamics Software
One of my longest-running projects is the development of open-source astrodynamics software. I try to make it high-quality, accessible, and understandable to enable others to build on it to solve their own challenges without having to reinvent the astrodynamics "wheel". This work started when I was beginning my graduate studies and found that there weren't easily accessible, easy-to-use astrodynamics software packages that I could both quickly integrate into my work and easily dig into the implementation details to understand how algorithms were implemented.
Selected Recent Work
AI Safety Red-Teaming (ASTPrompter)
Weakly supervised automated language model red-teaming to identify low-perplexity toxic prompts. Develops methods for automated failure discovery in large language models for safety-critical applications.
Failure Detection through World Modeling
Using world models to efficiently detect failures in autonomous systems, enabling robots and AI agents to learn robust behaviors from limited experience.
SecureForge
Finding and preventing vulnerabilities in LLM-generated code via prompt optimization. Even when frontier models are explicitly asked to write secure code, they still produce verifiable vulnerabilities 23% of the time.
Ground Station Optimization
Optimal ground station selection for low-Earth orbiting satellites. Uses integer programming to minimize cost while maximizing data downlink coverage for satellite constellations.
