I am a Ph.D. candidate in Computer Science at the University of Massachusetts Amherst, where I work with
Prof. Shlomo Zilberstein in the
Resource-Bounded Reasoning Lab.
My research focuses on agents built on large language models (LLMs) that negotiate, coordinate, and share resources with one another.
Their goals often only partly align, so each agent must decide what to say, whom to believe, and when to cooperate.
I'm interested in the following research areas:
- Trust and Reputation in Strategic Reasoning: Can agents learn whom to trust from past interactions? How should trust and reputation change what they propose, accept, and say?
- Social Agents and Strategic Communication: How do LLM agents negotiate, persuade, and build relationships when their goals only partly align? What social reasoning capabilities are required for effective collaboration with humans and other agents?
- Emergent Behavior of Multi-Agent Systems: What behaviors, such as collusion or favors that bypass the rules, emerge when agents interact over many rounds? When do these behaviors help a group, and when do they cause harm?
Previously, I studied how visualization supports differential privacy in practice.
Before my Ph.D., I worked on distributed constraint optimization with Prof. Md Mosaddek Khan at the Cognitive Agents and Interaction Lab (CAIL).
I also worked as a software engineer at TigerIT Bangladesh.
I received my B.Sc. in Computer Science and Engineering from the University of Dhaka.
arXiv 2026
