I’m a Postdoctoral Researcher in the Department of Computer and Information Science at the University of Pennsylvania, where I am advised by Lyle H. Ungar and Sharath C. Guntuku. Affiliated with the World Well-Being Project and the WorldBank. My research operates at the intersection of Computer Science and Cognitive Science, focusing on human-agent interaction, social norms and trust.

I earned my Ph.D. and M.S. in Cognitive Science from Brown University, advised by Bertram Malle, and my B.S. in Computer Science and Cognitive Science from the University of Michigan. My background also includes industry experience at Google’s Gemini Safety and Responsible AI team and the Honda Research Institute, where my research focused on human-centered intelligence.

Research Interests

How can we design AI chatbots that can teach, influence, and collaborate with people in sensitive, socially complex contexts,such as education and healthcare, while preserving human agency and well-being?

I employ a multi-disciplinary approach that spans human-robot interaction (HRI), large language models (LLMs), and computational social science (CSS). My current work is organized around four key themes:

• Norm Representation in Media: In collaboration with the WorldBank, I leverage LLMs to analyze how social and gender norms are represented in global media and develop tools to support more balanced cultural portrayals.
• Safety and Well-being: I develop interventions to ensure that AI interactions, particularly for vulnerable populations like adolescents,are psychologically appropriate and aligned with human values.
• Agency-Preserving Feedback Systems: I design conversational agents for educational and clinical training that provide reflective, “non-prescriptive” feedback. This approach prioritizes human autonomy by prompting self-reflection rather than delivering rigid corrections.
• Computational Modeling of Trust Dynamics: I use natural language processing and behavioral signal analysis to identify how trust is built and sustained during high-stakes human-professional encounters.

Not Research

• arthouse movies
• tennis
• bass guitar
• 🐴 English riding