Margarita Shebanova

Interview with Margarita Shebanova

How did you get in touch with Roche regarding your MA-thesis?

I did my masters programme in data science at Hochschule Luzern, and there is a special website where companies post their job opportunities for cooperation with students. Roche Diagnostics posted an opportunity about an internship at Roche in combination with the possibility of doing a Masters thesis. So I applied for this position. 

You have successfully completed a MA- thesis at Roche, could you tell us a bit, what is it about?

This Masters Thesis was for the demand forecasting project of the Global Supply Chain. And the goal of my Masters Thesis was to reveal the forecasting model that can deliver an accurate and robust forecast of demand. To achieve this goal, I compared different forecasting algorithms, namely statistical and deep learning models. As a result, the deep learning model called N-BEATS outperformed all the benchmark models on a given data set. Based on this success, now this model is being implemented in the demand forecasting toolset.

This is just amazing. How has your MA-thesis enhanced your professional life?

For me, it was the first advanced analytical project with real business data, and it also was a project in a very novelty field. By working at Roche, I learned a lot from people who work here because they shared with me best practices in the data science field. And now, I am continuing to work on this project at Roche as a contractor.

Why does it need theses like yours?

I think such a thesis can foster innovation at Roche because employees are usually quite busy with their day-to-day activities and don’t have enough time to investigate novel concepts. But students have this time and can do great research, sharing it with a broader audience afterwards. 

What would you advise future students?

I would advise them to collaborate with their project’s stakeholders as much as possible. And also to promote their results and present them to project management and other interested colleagues. It can create the opportunity for knowledge sharing to boost innovation and productivity in the company.

Where do you see yourself in 5 years

I see myself further developing in data science and advancing my analytical skills. By leveraging the technical expertise I will gain during the next five years, I also see myself taking more responsibility within the company and potentially leading data-related projects.

back to all success stories