Happy Choi
I am in my final year of Computer Science at the Hong Kong University of Science and Technology, with an Extended Major in Artificial Intelligence and minors in Mathematics and Big Data Technology. My final year project is on trust-aware LLM agents, and I have been doing undergraduate research on urban heat and building shadows since the summer of 2025.
About me
My full name is CHOI, Long Ho, and most people call me Happy. I am an undergraduate at HKUST, reading Computer Science with an Extended Major in Artificial Intelligence and a minor in both Mathematics and Big Data Technology.
For my final year project I am building TrustShop, an agent-orchestrated shopping assistant, supervised by Prof. Nevin Lianwen Zhang. The work covers trust-weighted review summarisation, a validated data acquisition pipeline, and the agent orchestration layer that routes open-ended queries. Separately, since June 2025 I have been working on the Heat Index Estimation project under the Undergraduate Research Opportunities Programme, supervised by Prof. FUNG, Jimmy Chi Hung and co-supervised by Dr. CHAN, Jimmy Wai Man and Dr. Tilman Leo Hohenberger. That project has grown from a Sky-View Factor study into a Hong Kong–wide temperature model, a 24-hour building shadow system, drone thermal imaging, and neural-network land cover classification.
In the summer of 2026 I interned as a Summer AI Intern at ESDlife, and in the summer of 2025 as an AI Programmer at the Centre for Intelligent Multidimensional Data Analytics. Both roles were mostly about building things end to end: scrapers, APIs, backend services and the analysis on top of them. The CV page has the full record, and Projects goes into the research in more detail.
What I work with
- AI & data
- Machine learning, deep learning, LLMs and AI agents, NLP, image processing, data mining, reinforcement learning
- Programming
- C++, Java, Python, R, JavaScript
- Web & systems
- Node.js, React Native, PostgreSQL, Linux
- Geospatial
- QGIS, ArcGIS, drone thermal imaging, solar irradiance and building shadow modelling
Contact
If you would like to talk about research, code, or anything else, email is the easiest way to reach me and I will do my best to reply.