We build reliable solutions that reinvent agriculture with the capabilities of AI and robotics. Our core system AmigaOS is modular, designed as adaptable building blocks for fully autonomous agriculture, and enables our hardware and software products to support a suite of robotic farming applications.

I joined Companion at seed stage to bring a new product category to market: a smart device using AI to play training games with dogs. I led SW bring-up for our new HW revision designed for manufacturing at scale.

I led efforts to improve coverage and quality of administrative audit logging for Insider Risk across Google. I worked similarly to expand Access Transparency, an enterprise security product, across Google Cloud Platform offerings through improvements to our internal onboarding process.

I led Daydream Data Infrastructure, providing Smartphone AR product teams with workflows for metrics and experimentation. My teams supported Lens ("search what you see") and ARCore (augmented reality SDK for mobile apps).

I managed the Tango Production / Metrics team responsible for internal infrastructure anddeveloper tools. Tango was the predecessor project to ARCore. I provided evaluation frameworks for computer vision algorithm quality and build/test/release infra for a globally distributed team of engineers, vendors and partners.

As employee #4, I led infrastructure for Project Tango from its early beginnings. ATAP was acquired into Google, and my primary work is covered above. I also worked on whatever else was needed, including device bringup, camera calibration, core architecture, and visualization tools.

I prototyped product concepts for an early-stage startup.• Fit recommendations for online clothes shopping: I developed tools for statistical analysis of human body shapes.• Web photo collage generation: I prototyped client-side algorithms for image effects.

I accelerated computer vision and robotics R&D in academia and industry through open sourcecontributions to OpenCV, Robot Operating System (ROS) and Point Cloud Library (PCL).• I authored the ROS “image pipeline” from capture to 3D reconstruction.• I designed OpenCV APIs and implemented algorithms for 2D features and object detection.• I integrated vision on the PR2 robot for successful autonomous tech demonstrations: plugging itself in, planning shots on a pool table, and recognizing its location.

I completed two summer internships and worked part-time during the school year, implementingCV algorithms and robot operations tools.• I implemented Interest Point and High Dynamic Range modules for the NASA Vision Workbench, a modern C++ computer vision library.• I developed terrain reconstruction and footfall planning tools for the ATHLETE six-legged lunar rover.