Research Experience

PAIR (People, AI & Robotics) Lab @ Georgia Tech

Student Researcher
Aug. 2026 - Present

Improving robotic foundation model architectures and training recipes for imitation learning using pre-trained foundation models and mixing real-world and simulation data.

Georgia Tech Automated Algorithm Design Team

Student Researcher — First-Author Publication at doi.org/10.48550/arXiv.2605.15649
Oct. 2024 – May 2026

Led research on generalizable performance predictors for evaluating machine learning models during Neural Architecture Search, using LLMs to embed candidate model code into numerical encodings for a lightweight fitness regression model. Published first-author work at GECCO 2026 on building predictors that generalize across model families without costly retooling during evolutionary search.

Georgia Tech Research Institute

Graduate Machine Learning Research Assistant
Aug. 2025 – May 2026

Developed surrogate machine learning models to approximate high-fidelity physics simulations and built a data pipeline for generating large-scale timeseries samples of simulated sensor signals from varied starting conditions. Applied physics-informed neural networks with transformer architectures for signal forecasting and used AutoML to tune signal processing models for classifying sensor data features.

Georgia Tech Research Institute

Machine Learning Research Assistant
Jan. 2025 – April 2025

Improved an LLM-guided evolutionary algorithm to enhance YOLO object detection models through neural architecture search. Cut runtime for model refinement and training by 22% by running LLM inference through a central server with a ZeroMQ message queue.