I’m an undergraduate in Computer Science at the University of Hong Kong, with a background in finance and a long-standing interest in how intelligent systems can actually help people work.
Over the past few years I’ve moved from general “AI interest” into a more specific focus on agent systems and context engineering – how to design tools, memory, retrieval, and protocols so that LLM-based agents behave reliably in realistic environments. This interest has been shaped by research on web-world modeling and tabular reasoning (TABULA-R² with JHU) and by my internship at ByteDance’s Volcano Engine, where I worked on tool-calling evaluation and an on-call assistant for SRE teams.
Most of my projects sit at the boundary between systems and intelligence: building executable models of the web, defining PLAN/END-style protocols for reasoning, and instrumenting agent behavior so that failures are observable and fixable. Going forward, I want to keep exploring how to turn large models into dependable collaborators for decision-making in data-rich domains.
Research Interests:
Full Resume in PDF.