I am an MPhil student in Electronic and Computer Engineering at HKUST. Before HKUST, I studied Computer Science at CUHK.
My current work sits around LLM agents, reinforcement learning, and applications in clinical decision support and robotic systems.
- Agent and tool-use systems: benchmarks, reusable skills, retrieval workflows, and practical agent harnesses.
- Healthcare AI: multi-agent and RAG-based workflows for medical-record evidence extraction, diagnostic reasoning.
- Robotics and control: reinforcement learning, robotic inspection planning and Sim2Real-style problems.
- Research software: small tools that make papers, experiments, and domain workflows easier to reproduce.
- ToolBudgetBench: a benchmark and demo harness for measuring LLM agent tool-context budget tradeoffs.
- DrugClaw: an OpenClaw-native drug discovery assistant for literature analysis, ADMET prediction, and virtual screening.
- investor-panel-skill: a skill that lets a company be reviewed from multiple investor perspectives.
- hk-ipo-skill: a skill for Hong Kong IPO information gathering, cross-checking, scoring, and review.
You can find my academic homepage here: li-jialu.github.io