About Me

Cyrus Asgari

I like solving complex technical challenges that have real-world impact. I like to wear hats as an engineer, data scientist, product manager, and leader.

I spent four years at Harvard where I earned a Bachelors degree in Applied Math and Master's degree in Computational Science & Engineering. I took (a lot of) classes focused on machine learning and optimization, while also indulging in economics (earning a secondary) as well as other cool topics like linguistics. I decided to spend some time over at MIT, enrolling in really cool graduate coursework in the fields of computer vision and sensorimotor learning for sequential decision making.

Outside of classes I was heavily involved in robotics, serving as the Co-President of the Harvard Undergraduate Robotics Club for two years and working on some really cool projects. I also was involved with the Product Lab group, serving as a product team leader and working with companies like Strava and Oracle. I gave some insights which appeared useful and were incorporated by our partners. I spent time researching reinforcement learning and methods of uncertainty quantification under Prof. Koumoutsakos, developing uncertainty-aware policy gradient methods for improved sample efficiency and robustness. I also dedicated a large chunk of my free time to teaching. I taught courses in Nonlinear Dynamical Systems, Applied Linear Algebra and Big Data, and Systems Development. I developed a love for teaching!

I also spent some time in various industry roles. Most recently, I have worked as a machine learning engineer at LinkedIn within their ads org. I worked on cross-domain graph neural networks to improve CTR. Prior to that, I worked at Biospectal, a startup developing an app to measure blood pressure using a phone camera. I developed a system for monitoring longitudinal data and detecting any aberrations. I also spent a summer at Amazon as a software engineer, developing a skinnable video player to expand their adaptive bitrate streaming tech to other orgs like Twitch or Music.