I'm the Director of Perception at Argo AI.
Before Argo AI, I led computer vision and computational performance teams at Alphabet. We created (among other things) the depth sensor powering Face Unlock on the Pixel 4. Our team led this effort from the ground-up, with deep involvement across the full stack - everything from silicon, to hardware component selection and design, to algorithmic invention and productization under considerable constraints. Additionally, I spent some of my individual technical time developing critical technology pieces in the ARCore Depth API.
Before joining Alphabet, I spent many years working on surface reconstruction for Microsoft HoloLens, particularly around high-performance algorithm development, meshing, and system architecture for the Spatial Mapping platform. Prior to my involvement in HoloLens, I worked on application anti-tampering and integrity in Windows 8.1 and Windows Phone 8.1.
Going yet further back in time, I attended Ohio State University, where I graduated magna cum laude with Honors in Engineering and an Honors Minor in Math from the Computer Science & Engineering Department. I worked in Dr. James Davis's Computer Vision Lab as an undergraduate research assistant, interned at the Ohio Supercomputer Center, Chemical Abstracts Service, and Microsoft.
In my off-time, such that it is, I enjoy home electronics/automation projects and hiking. Some side projects of mine include a home-based virtual assistant system with Phillips Hue integration and voice control, a lamp that displays the outside temperature by color, wirelessly-controlled LED lighting in my desk, and a couple of ESP8266 microcontrollers that I'm fiddling with while figuring out the next project.
Implemented improvements to the modern app platform confidence model (AppContainer), designed and developed anti-tampering protection for Windows 8.1 applications, contributed to the web application platform for Windows 8 Store apps written in HTML5/JS. Additional confidential feature work.
Responsible for creation of all AppModel trust components for the Apps on SD feature.
One of a small team of engineers building the realtime 3D reconstruction algorithms and logic. Author of all platform meshing code, Spatial Mapping system architecture, simplification and level of detail system, collision detection, depth image processing, debug visualization, caching logic, deterministic playback, work scheduling, flat region detection, and a large majority of the vectorization in the runtime. Performance and architecture owner for Spatial Mapping; drove relationships with silicon, systems, and performance teams, mentored junior engineers and led prototyping efforts.
Performance optimizations ranging between 500% and 2800% to core controller tracking algorithms.
Led the team that built the depth sensor powering Face Unlock on the Pixel 4.