Was a software developer at the UW Intelligent Networks Laboratory, where I worked on Graphitti Workbench, a graphical tool for managing eScience simulations. The tool was written in Java and used JavaFX.My contribution was updating the tool to work with the laboratory’s new simulator, Graphitti, from the old simulator BrainGrid, getting it closer to becoming a production-ready tool.I also wrote internal developer documentation in Markdown documenting the areas of the project I worked on to better onramp future contributors.

For my CodeLabs internship I worked with my mentor Yang Wei and a couple other students to re-train Pafnucy, an open-source 3D convolutional neural network (CNN) that predicts the binding affinity for protein-ligand complexes. To showcase its usage, we decided to use Pafnucy to predict the affinity (or effectiveness) of potential therapeutics for COVID-19. I personally:• Utilized pandas to process nearly 18,000 protein-ligand complexes from a database and split it into sets required for training the model.• Created a Jupyter notebook which uses Seaborn, matplotlib and other various python scientific computing packages to assess the performance of the model after training.• Identified potential ligands which can nullify the activity of corona virus associated proteins based on predicted binding affinity.

Was one of the programmers on my high school robotics team, becoming a lead my senior year. I was also an officer, acting as the team secretary my sophomore year.I started out building small parts in the codebase here and there. Most of my work in my last two years involved computer vision, developing efficient pipelines that would be used to help the robot automate complex tasks, such as interacting with certain elements on the field, or scoring game pieces.I would also teach new students about programming, computer vision, and how it can be used in robotics.