background

I am from Severodvinsk, the city in the north of Russia near Arkhangelsk. Near the end of school, I decided to connect my life with Computer Science, that's why I applied to the Saint Petersburg State University of Aerospace Instrumentation for Information Security. In my second year, I understood that data protection is not so interesting for me and I started to work as a software engineer until the graduation.

After two years of working as a software engineer, I understood that I want to work in the field closer to math and science. At the same time, I found Data Science and was excited about it. So one year I spend on Coursera and Stepic courses but realized that my progress is going slow and I need more profound knowledge. And I started to find a way to get them.

This year we applied to NTI Olympiad at Skoltech (I have heard about it because my university has a joint program). There I used the opportunities to get an interview with the admission board of the Skoltech Information Science and Technology program. After we talked with professors and I saw the campus, I understood that is an excellent opportunity and I should use it.

In the end of competition, I was accepted to the program, and we took third place!

I decided to be a part of the Joint Laboratory of Applied Research Skoltech-Sber (LARSS) because it is an excellent way to work while you are a student. Also, in the lab's presentation I saw several interesting research topics for me.

In LARSS I work on several topics such as predicting user's lifestyle changes, neural networks for tabular data and several others.

My summer industrial immersion was in Sber, and after it I started working in the Risk R&D department.

Skoltech and the Information Science and Technology program, particularly, the place where big amounts of brilliant people surround you, motivate you to dive deeper into your exciting sphere. This program could give you modern technologies and skills to go into the industry right now.

Now I am working on my thesis project "Practical applying of transformer for financial data". Several years ago, transformer architecture became SOTA for natural language process tasks such as machine translation, summarization, etc. Therefore, we try to adopt this architecture for the different domain - financial data. We want to investigate the application of this model to help bank solve different type of tasks on sequential data more efficient.

In the near future I plan to improve my data science skills, become an expert in my current field as an engineer, and influence solving global machine learning problems. Once this goal is fulfilled, I may have to think about trying myself as a manager.