• Designed and implemented a sparse spatial data-structure for parallelized convolutions using intel’s tbb, speeding up support generation by 30%.

• Created an in-house FEA simulation for arbitrary models using Eigen’s iterative and numerical solvers.

• Sped up STL loading by up to 700% in the cases of high poly models through novel parallelized brokenness detection to create one of the fastest STL loaders on the market.

• Designed and launched an internal tool for monitoring the status of problems in various Google Coding Competitions to improve the problem creation workflow for developers.

• Implemented the backend services including a cache for sending RPCs in C++ which involved designing a complex system to communicate between different internal services.