Visiting Scholar, The Hong Kong University of Science and Technology
Mingchen Li is a PhD candidate in Civil Engineering at The Hong Kong University of Science and Technology and currently a visiting scholar at UC Berkeley. His research focuses on developing AI-driven tools for querying and analyzing building semantic data. By combining large language models (LLMs) with knowledge graphs, he aims to make technical building information more accessible through natural language.
His current work involves designing retrieval-augmented generation (RAG) pipelines to generate accurate SPARQL queries from user input, applied to domains such as HVAC system modeling, energy data integration, and building performance analytics. In addition to academic research, Mingchen has practical experience in BIM design, parametric modeling, and Python-based building simulation tools.