Hugh Flournoy Van Deventer V
Data Science & AI Research
Incoming Master's Data Science student at Harvard SEAS, following my studies in Mathematics, Physics, and Computer Science at the University of Michigan. My research interest centers on how complex behaviors arise from simple rules across diverse domains, including physics, mathematics, and AI. Understanding this emergent complexity, through approaches like mechanistic interpretability in AI, is crucial, as it directly impacts our ability to predict, control, and ultimately ensure increasingly powerful AI systems operate safely and beneficially for humanity. I am excited to start deeply engaging in this research at Harvard.
Currently working as a Data Science Fellow at the UM Center for Academic Innovation. Incoming AI/ML Intern at Honeywell.
Research & Publications
From Interests to Insights: An LLM Approach to Course Recommendations Using Natural Language Queries
We introduce a novel course recommendation system that combines large language models with semantic search to bridge natural language queries and course descriptions. Our two-stage approach generates idealized course descriptions from student interests, then uses embedding similarity to identify relevant courses. Empirical evaluation demonstrates the system's ability to provide contextually appropriate recommendations with explanatory rationales.