I am a researcher in the field of artificial intelligence, focusing on multi-agent systems and strategic reasoning. Currently, I am a postdoctoral research associate at the University of Southern California, working with Milind Tambe and the Teamcore research group. I received my Ph.D. from the University of Michigan, where I was advised by Michael Wellman. During my time at Michigan I was active as a fellow of the STIET program, which provides interdisciplinary training and brings together researchers from diverse disciplines with shared interests in incentive-centered design.

My primary research interest is developing new computational tools for strategic reasoning in complex multi-agent systems. Understanding how to make good decisions in the presence of other intelligent agents (cooperative or adversarial) has always been an important problem. In today's world, it is a crucial challenge for computer science, as computer systems become increasingly autonomous and interconnected, and technology is used to enable new forms of human interaction at a staggering rate. Recent progress in multi-agent systems, game theory, mechanism design, and related areas has produced many new new tools for modeling and analyzing systems with multiple decision makers. However, applying these paradigms to real-world applications presents an array of new challenges. We need new solution algorithms that are computationally scalable to systems with hundreds or thousands of agents, and extremely large strategy spaces. Model uncertainty is ubiquitous when describing complex, dynamic scenarios, necessitating new methodologies for building models from empirical evidence, and new solution techniques for noisy, incomplete domain models. Finally, we must develop a greater understanding of how to reason about agents with bounded rationality, including humans.

I consider applications to be a major component of my research agenda. Recently, I have begun work on two new projects with colleagues at USC. The first is a security patrolling domain, where we are using game-theoretic methods to create randomized schedules for the Federal Air Marshals. In the second project I am developing distributed optimization tools for team formation, driven primarily by the requirements of disaster response scenarios. I have been heavily involved in the Trading Agent Competition for many years, starting as an undergraduate when I assisted with running the tournament. I am one of the lead designers of the very successful Deep Maize agent, which won the tournament in 2008 and has consistently been among the top few performers in every tournament to date.

Research Interests: artificial intelligence, multi-agent systems, computational game theory, machine learning, trading agents, supply chain management, security applications, distributed optimization, mechanism design, behavioral game theory, team and coalition formation