I am a research scientist, engineering leader, and software architect.

I was most recently a principal research scientist and warehouse flow engineering lead at inVia Robotics. My day-to-day work involved leading a team that thought about how robots, people, and goods should flow through warehouses. Working at a start-up, I also got to play a role in interacting with customers, whether it be to help guide product development, or assist in the pre-sales process.

Previously, I was as a Ph.D. student studying computer science at the University of Southern California, advised by Laurent Itti. I focused on robotic perception, participating in the Darpa Robotics Challenge, before pivoting to working on feedback in neural networks. My thesis, "Learning Invariant Features in Modulatory Neural Networks Through Conflict and Ambiguity", centered around a new, unsupervised learning rule, called "conflict learning", in conjunction with a mechanism to reduce ambiguity in neural networks. I was inspired by a process in the visual system known as "border ownership", where neurons selectively respond to the "inside" of an object.

Although my recent work has focused largely on multi-agent pathfinding, the intersection of biologically inspired approaches with artificial intelligence remains a strong passion. I am especially interested in the impacts and benefits feedback can have on neural networks.

Outside of work I'm an avid hiker, climber, reader, and media-enjoyer. The picture above is en-route to a popular but still very worthwhile destination in Norway, Trolltunga, taken on a clear snowy day in October.