Visit the project on BitBucket or watch video 1 and video 2, which demonstrate the algorithms. Used to quickly find realizations of rigid, 2D bar-joint graphs. This was my personal project in the PhD program, for which I developed the theory and algorithm, which is suitable for industry CAD software. I independently coded and architected the software that implements the algorithm and provides a graphical interface. See publications for related papers.
Visit the project on BitBucket. Used to explore the assembly landscape of molecules (and other physical structures.) I led the redesign of the UI and the refactorization and restructuring of the codebase, which has been in development for ~10 years with ~5 active contributors, allowing for accelerated development with undergraduate students. I also helped write the user guide and feature summary (see publications.)
Machine Learning Project
Coded many ML algorithms throughout the semester in Python, outputting graphs via Matplotlib. This culminated in a large project (with one partner) that compared and contrasted multiple forms of unsupervised nonlinear dimensionality reduction and manifold learning algorithms on a variety of data sets.
An independent project. Implemented in C++ from the ground up (using only minimal windowing and asset loading libraries.) Helped develop a deep understanding of the modern OpenGL programmable pipeline (i.e. using shaders). Features several sophisticated software architecture patterns, notably the entire system backbone is built on an event queue which allowed me to use multi-threading for individual systems. Also features Lua scripting integration.
Visit on GitHub. An online app made as an entry for the 2016 Riot Games API Challenge, by myself and one partner. Features: dynamic website powered by Ruby on Rails, attractive UI, large PostgreSQL database (~90k users), daily tasks for updating database via calls to Riot Games API, and more.
Note: No longer being hosted.
A slick and beautiful website, right? No doubt the envy of other professionals. Built from scratch. Deployed through GitHub pages. See source here.