Hi, I'm Juncheng Shi๏ผ็ณ้งๆพ๏ผ, currently a Research Assistant at HKUST(GZ), working on Embodied AI.
My core goal is to build intelligent systems that can perceive, reason, learn, make decisions, and act in real-world environments.
Toward this goal, I mainly focus on RL, Agents, and LLMs, exploring how interaction-based learning, planning, tool use, memory, multi-step decision-making, language understanding, reasoning, and task planning can work together to support embodied intelligent systems.
Through these directions, I hope to enable intelligent agents to perceive, learn, make decisions, and complete complex actions in real physical environments, turning ideas into practical technologies that can truly serve people.
- Embodied AI and intelligent systems
- Reinforcement Learning
- Agent systems
- Large Language Models
- AI integration with physical / simulated environments
- Vibe Coding and emerging development workflows
- Build solid foundations in RL, Agents, and LLMs
- Explore embodied intelligence through simulation and practical projects
- Develop reproducible experiments and well-documented repositories
- Improve engineering skills for building reliable AI systems
- Keep learning, building, and shipping
| Project | Description | ๆๆ |
|---|---|---|
| MiniCode | Lightweight terminal coding assistant with Claude Code-like workflow and TUI. | |
| Shushu Internship Tool | AI-powered internship assistant that transforms job descriptions into project plans, resumes, and interview preparation materials. |
- GitHub: @GateJustice
- Email:
GateJustice0815@gmail.com