About me

I am an applied research leader in machine learning and computer vision. I have repeatedly guided teams out of dysfunction and to realize a well-oiled, high-performing and enjoyable work environment, all while building deep tech products. Further, I enjoy making connections in applied research topics related to machine learning and deep learning, and relating them to product visions. In 2022, I was named an Emerging Leader in AI in Canada, before I moved (back) to the United States.

I am a Staff Software Engineer with the Perception team at Stack. Here, I develop deep learning based systems that enable our autnomous trucks to sense the world around them. Previously, I was Head of Machine Learning at Geopipe in NYC where I led the team responsible for developing deep learning models to perform 3D reconstruction of digital twin cities from orthophotos and LiDAR data. Prior to that, I was Team Lead and Scientist of the Machine Learning team at Invision AI in Toronto where I supervised and grew a machine learning team to develop algorithms for use in smart infrastructure technologies such as object detection, knowledge distillation, federated learning and multitask loss optimization among others. Prior to that, as a Computer Vision Scientist I worked on top-down 3D human pose estimation as well as fully-differentiable bottom-up pose parsing using transformer networks at wrnch AI in Montréal which was acquired by Hinge Health. I have also contributed to and consulted in applied and foundational topics ranging semi-supervised learning, knowledge distillation, shape estimation, neural inverse kinematics, video anomaly detection, multi-modal learning, video action recognition, person re-identification and adversarial learning.

I have a Ph.D. from the Computer Science department at North Carolina State University and my dissertation was titled “Anomaly Detection in Videos”. I was advised by Ranga Raju Vatsavai and also worked actively with Mike Jones at Mitsubishi Electric Research Laboratories.