Kenny Wongchamcharoen
Industrial Engineering & Operations Research
University of California, Berkeley
Berkeley, California
I’m a senior at UC Berkeley, majoring in Industrial Engineering & Operations Research and minoring in Data Science and Mathematics. My research lies at the intersection of operations management, AI, and data-driven decision-making.
I study how to design large-scale service systems under strategic interactions and information asymmetry, with applications in transportation and public policy. I also evaluate the capabilities and limits of AI in operations and study how to best integrate intelligent agents into workflows to optimize operational performance. My work draws on operations research, data science, and applied economics. More about my research interests →
At Berkeley, I am privileged to work with and be advised by Chiwei Yan in IEOR and Abhishek Nagaraj at the Data Innovation & AI Lab (DIAL). I spent summer 2025 as a visiting research fellow with Paul Glasserman at Columbia Business School’s Decisions, Risk, and Operations (DRO) division. I also collaborate with Phillip Kerger on designing and teaching various courses at Berkeley.
Latest news
All updatesI'll present On-Off Systems with Strategic Customers at the MSOM Invited Session of INFORMS in San Francisco. I'd love to connect!
CentaurBench was accepted at the NeurIPS 2026 Workshop on Agentic AI Benchmarks and Applications for Enterprise Tasks (AABA4ET)!
I’ll be presenting CentaurBench at the Wharton Generative AI & Business Conference on September 9–10, 2026.
Awarded the Tau Beta Pi Scholarship for academic achievement, extracurricular activities, and promise of contributions to the engineering profession.
Selected research
All researchCentaurBench: Benchmarking LLM Capabilities on Augmenting vs. Automating Real-World Work Tasks
A model's ability to automate a task is distinct from its ability to assist another agent. Effective human–AI and AI–AI workflows require evaluating models for the roles they play, beyond their standalone performance.
Do Large Language Models (LLMs) Understand Chronology?
Testing the limits of chronological reasoning in LLMs, with implications for look-ahead bias in forecasting.