We introduced a Multi-Agent Path Finding framework that allows us to minimize the coordination needed between agents. The key idea behind this method is that we search for paths within the "Space-Level" state space, rather than in the Space-Time state space. This allows us to minimize the coordination between agents by minimizing the number of levels in the final solution because each agent can move freely, at an arbitrary velocity in each level. We have shown a 20-50% reduction in coordination when compared to the current state-of-the-art method.

In the summer of 2022, I worked as a Software Engineering Intern at NVIDIA as part of their autonomous vehicles team. While at NVIDIA, I was tasked with removing security vulnerabilities in the code to ensure that we could keep our team's software safe and reliable for our customers. Throughout my time at NVIDIA, I removed over 1,000 security violations from various parts of our team's software. This required me to reach out to many different co-workers to ensure that each solution worked for everyone and did not introduce any errors into the code. I was also able to enforce automatic violation detection, ensuring that these violations could no longer be introduced into our software.

In the summer of 2023, I was tasked with updating a sensor visualization tool to allow for sensor interfacing and data visualization to occur on separate machines. I implemented communication between separate processes through sockets to offload computational overhead from the on-car computer.

In the summer of 2021, I had the opportunity to intern at the NASA Jet Propulsion Laboratory. Here, I was part of a team that was developing small-scale flight software. My task was to create a tool that would automatically create components of this software for the developer, greatly reducing the amount of time it would take developers to create a project using this flight software. Throughout this task, I reached out to many users, ensuring that my tool would be accessible and beneficial to all. In addition to creating this tool, I also worked on implementing continuous integration. This helped to ensure that any new software updates would not cause any issues or introduce any backward compatibility problems. My time at the Jet Propulsion Laboratory helped me greatly develop my software engineering skills, as well as my communication and collaboration skills. 

This is a NASA-supported project, led by Dr. Red Whittaker, with the objective of designing and constructing an autonomous lunar rover set to explore the moon’s pole in 2023. While I was given the opportunity to work in many different fields within this research, I spent the bulk of my time developing perception software. I successfully developed a perception filter that removed unfit frames from our perception pipeline. Here, I was presented with countless obstacles, as developing a perception system for lunar use is drastically different than it would be here on Earth. Applying my software engineering skills to these challenges has allowed me to grow and adapt to the many challenges faced when developing any software.

Worked with a team developing a system capable of generating an optimal race line for an autonomous vehicle around a given track. This involved researching the use of constrained optimization to create the most efficient solution, while considering the physical limitations of the vehicle.