Architect and software lead for an end-to-end product that leverages machine learning to transform Sonar data into 3D worlds which were automatically rendered in Unreal Engine, Gazebo Isaac Sim, or Gazebo (python React, javascript, electron, CUDA, pytorch). I also wrote ROS 2 for various underwater robotics products, and contributed to Makai's flagship commercial software (Makai Plan, Makai Plan Pro Makai Lay, in C#/C++).

Developed and implemented AI agents to use Generative AI, OCR, and Email Webhooks tools, to create an to efficiently scrape insurance data from diverse PDF and image sources, and upload them to our database.

Developed a neural network using PyTorch to predict binding energies of nucleic acid-protein pairs.

This was part of a multidisciplinary team at Oceanit, focused on creating a Covid-19 detection test.

Created novel quantum lattice gas algorithms for the simulation of the non-Abelian physics of spin-2 BECs, and of 2 particle systems. The entire algorithm was written CUDA. This included writing sophisticated 3D visualization to aid with the understanding of the entire 5 dimensional spin manifold of the spin-2 BECs