I'm a Ph.D. student in Applied Mathematics at UC Riverside, working at the intersection of machine learning, LLM agents, and applied mathematics, with an emphasis on reliable AI systems, rigorous evaluation, and mathematical frameworks for complex AI-driven systems.
I design benchmarking frameworks and adversarial evaluation environments for LLM agents, from automated benchmark generation to isolated execution pipelines and cross-model evaluation. My work on Backtrader-Bench, a benchmark for LLM agents on algorithmic trading, was accepted to the FinLLM Workshop at IJCAI 2026.
I also study multi-agent AI competition through dynamical systems and evolutionary game theory, and work on scientific machine learning with neural operators (In-Context Operator Networks) and Gaussian process methods for PDEs and biomedical systems. I'm drawn to problems that demand both mathematical depth and end-to-end engineering.
I've conducted research at UC Riverside, Lawrence Berkeley National Laboratory, Argonne National Laboratory, and Voaige, spanning AI agents, AI for science, computer vision, and finance, and I teach mathematics at UCR.
Beyond building and evaluating AI systems, I'm increasingly interested in how they reshape labor markets, economic incentives, and society as they grow more capable, and I'm expanding toward reinforcement learning, agent training, and scalable evaluation.
When I'm not doing research, you'll find me crocheting, hiking, or on the slopes snowboarding.
