Developed efficient method of keyword semantic similarity with text embeddings and an approximate nearest-neighbors algorithm, improving similarity identification by over 200%.Created data pipelines in Spark that processes billions of samples for production-scale model training and evaluation.

Created survey web application that allowed users to view research projects they were responsible for and stored survey responses in a backend list that was exportable to Excel. Wrote Python scripts that compared and detected missing research reports between catalog and server to ensure smooth and successful migration to the new database.