About
I am a Master's student at Harvard studying Data Science with a background in Math, Physics, and Computer Science from the University of Michigan. My research interests in AI are focused on interpretability, safety, and the science of deep learning.
At Harvard, I am currently working on masked diffusion models. I'm also a Winter 2025 FIG Fellow, working with Eleni Angelou on how to reliably evaluate activation-based interpretability methods.
At Michigan, I was a Data Science Fellow at the Center for Academic Innovation, where I developed ExploreBlue, a course recommendation system using LLMs and semantic search under the guidance of Prof. August Evrard and Dr. Mark Mills. I also served as a peer advisor for the Physics Department.
This summer, I worked as an AI/ML Engineering Intern at Honeywell, where I worked on AI and classical image processing techniques to automatically extract building layouts from floor plans and generate cost estimates for corporate facility projects. Previously, I was an ML Research Intern at Michigan Tech Research Institute, where I worked on ML methods for accelerating inverse problem solvers under the guidance of Dr. Joel LeBlanc, and at Neurabuild in South Africa, where I developed ML solutions for astronomical site assessment and satellite tracking for the Astrosite project under the guidance of Prof. Gregory Cohen.
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.