Cooperation
How can agents learn shared strategies when observations, actions and rewards are distributed?
Research
I study how autonomous agents learn, coordinate and adapt—especially when information is incomplete and the world changes around them.
Core themes
The long-term goal is to make groups of intelligent agents more capable, understandable and dependable.
How can agents learn shared strategies when observations, actions and rewards are distributed?
How can behaviour remain effective with unfamiliar teammates, opponents and environments?
How should agent systems be evaluated when coordination fails or conditions move beyond training?
Approach
These three ideas provide a simple lens for connecting algorithms, experiments and real-world intelligent systems.
Discover effective behaviour from interaction and feedback.
Share goals and act coherently across multiple decision-makers.
Respond to changes without rebuilding the system from scratch.