Professor Haohan Wang recently joined the Beckman Institute.

“Beckman’s culture of deep interdisciplinarity and methodological rigor aligns closely with how I think about AI for science,” he said “I am particularly excited to engage with researchers who treat AI not as a black-box tool, but as a system that must be understood, stress-tested, and co-designed with scientific objectives in mind.”

Haohan’s research centers on principled multi-agent AI systems that move beyond prompt-level orchestration toward controllable, optimizable and testable techniques and architectures used to manage and orchestrate the actions of these systems. At Beckman, this work connects closely with researchers in intelligent systems, computational neuroscience, imaging and data-driven life sciences. It allows for collaborations between AI methodology and scientific applications. He intends to expand collaborations that combine foundation models, agent architectures and domain-specific scientific pipelines.

More broadly, he’s interested in researching agentic and multi-agent AI systems for scientific discovery, with a particular emphasis on computational biology and trustworthy AI. His work bridges machine learning theory, systems and real scientific workflows.