Julia is an AI Research Engineer at Dryft, building at the intersection of agents and mathematical optimization.

Julia graduated early from MIT with a double major in Artificial Intelligence & Decision-Making and Mathematics. At the MIT Biomimetic Robotics Lab, she worked on the MIT Mini Cheetah and Humanoid robots and quadcopters. Her research focused on reinforcement learning and neural network architectures for hybrid learning- and model-based controllers. As chair of the MIT Schwarzman College of Computing’s undergraduate board, she built NerdXing: MIT’s first website for exploring students’ course histories and pathways for interdisciplinary studies. As President of MIT IEEE/ACM, she led the largest EECS club on campus, guiding a 20 person team to hold 25+ events for 1,000+ attendees annually.

Before Dryft, Julia designed performance benchmarks for vision language models and statistical metrics to measure fine-tuning data quality for large language models at NVIDIA. She also developed learning-based autonomous navigation for helicopters at Lockheed Martin. At NASA, she developed machine learning models for forecasting the spread of vector-borne disease and presented her work at the American Geophysical Union. Her work has earned her national recognition as an Astronaut Scholar, US Presidential Scholar, and three-time AIAA scholar, among others. She is also an internationally-recognized classical pianist, avid social dancer, and philosophy of science enthusiast.