After I got my bachelor's degree in physics at Moscow State University I have chosen to continue my academic way in Machine Learning and Data Science. This area has excited me a lot then and excites me even more now. I have heard of Skoltech as an innovative university that provides students with an up-to-date theoretical basis and practical skills. Also, I considered Skoltech because of its professors and high-level education.

The journey was a challenge. I started my industrial way as an analyst at Yandex Market. Then I worked on recommender systems. These jobs helped me a lot to dive into industrial ML, the approaches that are used in big tech companies. I also participated in numerous competitions. Two of them were devoted to the problem I work on now (Lyft Motion Prediction for Autonomous Vehicles and Waymo Motion Prediction Challenge). Both times my team has shown a good result. Finally, the combination of academic background from Skoltech, industrial experience from Yandex, and the results from relevant competitions brought me to the Self-Driving Group.

Now I work as a Machine Learning Engineer at Yandex Self-Driving Group. More specifically I work on predicting the movements of other agents on the road. Surprisingly, it is exactly what I have been doing at Skoltech during the MSc program. A very direct application of university knowledge in the industry!

I believe it goes without saying that strong fundamental knowledge is crucial for solving challenging Machine Learning problems and pushing models to the limits. As for me: in addition to the fundamental courses, my MSc thesis was exactly about what I do now at work — predicting the behavior of other agents around the autonomous vehicle. It was my luck to work on this exciting problem at Skoltech and continue working on it in the industry.

My scientific advisor was Prof. Gonzalo Ferrer, the head of Mobile Robotics Lab. I was always keen on Self-Driving technologies and the lab and Prof. Ferrer provided me with the great opportunity to do research in this area. More specifically I worked on motion prediction of the agents (cars, pedestrians, etc) surrounding the autonomous vehicle. And once we know the intentions of those agents we can take our safe path. It is a very challenging problem because of the natural uncertainty of other agents' behavior. Moreover, the solution must be very reliable and fit all the constraints to provide a safe ride.

For those thinking of applying to Skoltech, I would say it is definitely worth a try. For students, I would recommend being proactive and taking every opportunity Skoltech professors and Skoltech community provide.