- Researched innovative pharmacogenetic correlations among more than 1,000 drugs and gene variants that are associated with changes in enzymatic activity and resulting phenotypes- Optimized a BERT language model to identify the latest indications for newly FDA-approved drugs, resulting in a 22% increase in the precision of the existing model

𝐓𝐞𝐱𝐭𝐀𝐭𝐭𝐚𝐜𝐤 𝐍𝐚𝐭𝐮𝐫𝐚𝐥 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐏𝐫𝐨𝐜𝐞𝐬𝐬𝐢𝐧𝐠- Contributed to the development of TextAttack, an open-source Python library designed for executing adversarial attacks against Natural Language Processing (NLP) language models.- Produced a step-by-step tutorial for launching adversarial attacks against a fine-tuned BERT language model

- Developed a plugin for the University of Virginia's School of Medicine AI Lab, which facilitates the quick preprocessing of Petri dish images. This includes noise subtraction, multistack registration, and thresholding

-Built out a full stack application for viewing virtualized tours of physical spaces using Angular and NodeJS-Leveraged Google Directions API to improve route fetching and navigation

-Self-conducted research on the application of machine learning algorithms to mammography datasets-Utilized a support vector machine trained on mammogram lesion numerical records, fine tuned using a genetic algorithm. Then implemented a convolutional neural network trained from a collection of digitized mammogram scans-Developed into web application with Python Flask and SQLite3-Presented at the Intel International Science & Engineering Fair, Virginia State Science & Engineering Fair, and the LCPS Regional Science & Engineering Fair-Awarded: Intel Computer Science Award, GoDaddy Open Source Award, Aerospace Corporation Award, Virginia Dental Association Award-Honorable Mention