Exploring novel approaches to clothes change re-id with deep learning. Responsibilities include generating new approaches to the clothes change re-id problem; training CNNs, ViTs, and efficient models with contrastive and traditional loss approaches; collecting and validating large datasets; and explaining network outputs and decisions.

Provided performance analytics for machine learning model predicting optimal distributions of marketing dollars; produced primary analytics for regional marketing campaign; helped to remedy important discrepancy in a performance metric affecting a large program; assisted in transitioning part of an analytics code base to python.

Software engineer for the Live ASL translator project. Transfer learning on lightweight CNNs to generate real time translations for over 100 ASL signs. Utilized cloud computing softwares including AWS sagemaker, cloud9, and s3.