Research

My research interests span multiple areas of computer engineering, including parallelism, security, computer architecture, and compilers.

If you're a current Midshipmen and interested in any of these projects, please contact me!

Generative AI for Warfighting

In the high-stakes environment of squad-level tactical combat, leaders are often overwhelmed by the sheer volume and disparate nature of information about, and capabilities of, friendly, enemy, and, increasingly, autonomous units. This project offers a transformative solution: employing generative AI as an intelligent assistant capable of managing and contextualizing battlefield data – including text-based communications, voice transmissions, and a myriad of sensor data, both locally and remotely generated.

Anti-Aimbot

Aimbots are commonly used to cheat in first-person shooter video games.  A program identifies a target on the screen and automatically aims and fires for the player.  In this project, we hope to limit the player's ability to cheat without falsely affecting good players, and we hope to limit the performance and security impacts of monitoring for cheaters.

I currently have an independent research student working on this project.

Edge Machine Learning with Deep Neural Networks

Machine learning using neural networks has exploded in popularity recently.  In this project, we analyze the runtime performance and energy costs associated with deep neural network models on edge processors.  We hope to leverage techniques like field-programmable gate arrays and computational sprinting to improve the performance of edge machine learning.

Memory Safety

Due to high performance costs, people have been unwilling to adopt full memory safety in unsafe langauges like C and C++. This has caused an arms race between security countermeasures and malicious attackers. I am interested in investigating a variety of approaches to solving this problem at the language, system, and hardware level.