Work & Experience

Researching machine learning and computer vision techniques for MRI imaging, advised by Dr. Olesya Melnichenko.

Engineered a video-language pipeline to automatically monitor ICU patient behavior from continuous footage, enabling real-time health assessment of 2,656 critically ill patients without manual intervention.

Built a single model that predicts user actions from unlabeled video across 9 distinct environments, achieving 85.2% accuracy and enabling zero-shot action understanding without labeled training data.

Leading strategic partnerships and external relations for Stanford's premier Extended Reality organization. Co-led launch of partnerships with Meta, NVIDIA, Amazon, and Snapchat. Organized immersive technology hackathons engaging 300+ participants in XR and Spatial Computing projects.

Drove development of CNN models to generate realistic 3D hair reconstructions across demographics, boosting accuracy to 96% and enabling deployment on lightweight devices. Built a highly robust extraction algorithm with OpenCV and SAM segmentation, capable of handling poor-quality images and achieving perfect segmentation on 98% of images.

Engineered an AR application on Snap Spectacles in collaboration with the Snapchat Prototyping Team, enabling analysis of 92% of the available field of vision. Created a process to transform video into accurate 3D environments by integrating SLAM and point-cloud segmentation models, thereby optimizing output by 44%.