Lead an engineering team to scale the capacity of motion planning stack for product needs. Define and productionize APIs and touch points for both classical and ML based motion planning algorithms. Areas of focus include cost definition, data quality pipelines, and overall onboard architecture of the planner stack.

Lead a team of roboticists and game dev engineers focusing on robot localization and control systems. Developing on both embedded and mobile platforms. Design and implementing complex robotics algorithms on low-level firmware while meeting computational and cost constraints. Design and implement high-level planning and decision making processes, including interface with core game mechanics.

I worked on the self driving car project. Specializing in path planning and decision making. Using machine learning techniques to correctly select control actions based on the robot's environment, obstacle intentions, and traffic regulations.

I worked on the Autonomous Robotic Manipulation (ARM) project sponsored by DARPA. I designed and implemented an algorithm that allows the manipulator to select the best grasp for objects based on a partial view of 3D point cloud data.

Principal software engineer at a 3-person start-up. Modiface uses state of the art facial feature localization research we developed at University of Toronto to perform automatic manipulation of images and makeup simulations.I oversaw all software design and development in the company, and worked with clients to customize our computer vision and image processing engine to meet their requirements and launch deadlines.While at Modiface, I also continued with research and development on the core image processing and facial feature localization technologies.