We are discussing the corporate master’s program at ITMO University and Napoleon IT

The education of a programmer is an eternal topic of discussion. There are adherents of the traditional paradigm of “unlearn at a university and go to work”, others choose courses on technology stacks that interest them, and still others try to learn the basics of programming in the fields. However, there are cases when a university and business work together, for example, by conducting internships or developing joint educational programs. Today we will discuss one of these programs — a corporate master’s program at ITMO University and a developer of high-load systems using cloud computing from Napoleon IT. “Machine learning engineering”.


Dmitry Botov

Associate Professor of the Faculty of Infocommunication Technologies at ITMO University, Ph.D.

Dmitry, this is the first experience of an online master’s program at ITMO. Why did you choose this format?

Our global goal unleash the potential of talented students to the Middle level, which means we must make their learning experience as close to real work tasks as possible. Research showsthat most employees evaluate the experience of remote work positively, and employers are ready to leave the hybrid format even after the removal of all restrictions, so I believe that remote work has become the new norm, and the student should be ready to interact remotely. For our part, we attract mentors from leading AI companies with the task of involving students in the working atmosphere and processes of our distributed teams, we implement training in the online project office format. As part of the program, we provide career support for talent from the beginning of training to the moment of employment, and I am sure that this approach will be more effective and personalized than the traditional format.

We are not creating a master’s program in the traditional sense, but an innovative progressive educational technology. Here, students do not study at the university and do internships, they work in the company. The most important idea – in an inverted approach to the training of engineers – is to take the best from the methods of engineering schools, combine it with interactive online learning technologies and best practices for working with interns in IT companies. ITMO is one of the leaders in the training of specialists in the field of advanced technologies aimed at the development of science and entrepreneurship, so it is absolutely logical that such a project was born here.

Who is the program aimed at?

We are looking for students with experience in mathematical analysis, who are well versed in linear algebra, probability theory and statistics, understand basic ML algorithms and neural networks, and can work in Python. Our program is suitable for existing IT professionals who want to move into the promising field of machine learning.

This is not another 2-year classic MA in artificial intelligence or an online machine learning course. There will be regular code reviews and implementation of the entire cycle of working with data and models, challenges to launch work products in real infrastructure and continuous communication with specialists in various areas of Data Science.


How will the training be structured? What projects will master students be able to work on in two years?

The training of ML engineers will take place in a real workflow in the format of a project office. The first year of study is an intensive training under the guidance of experienced mentors in classical methods of machine learning, computer vision technologies and natural language processing, as well as in the chosen specialization module.

There is only one criterion for success when passing the next project module, as in the industry – a working software service for all customer requirements for the deadline. The first year ends with an internship, a three-month internship in the project teams of one of the AI ​​partners, remotely or with relocation upon request.

In the second year, students will be able to implement the ideas of new products and technologies of artificial intelligence by choosing one of the non-classical formats for preparing a graduate work: develop an applied solution in project teams of industrial partners, get into a research group and conduct applied research in Data Science, participate in development of an open source project or upgrade your startup in the AI ​​accelerator.

We plan to deeply immerse students in projects that are not limited to the creation of models, so that they can go through all the stages of the work of a machine learning engineer: from collecting and labeling data with building pipelines for their delivery to training and deploying models with monitoring and evaluating the effectiveness of algorithms and ensembles models, optimization of training time and inference of deep neural networks, development of backend services and API integration with user interfaces.

During the training, undergraduates will be able to choose one of the specializations:

  • Data Mining in Retail and Marketing;

  • Computer Vision in Retail and Manufacturing;

  • Computer Vision in Security and Surveillance;

  • Conversational AI;

  • AI-generated media.

What positions in companies can graduates apply for?

As I said, the program is focused on training machine learning engineers. This profession is in high demand. In the global labor market, hundreds of thousands of such specialists are already required, only in the USA there are up to 50 thousand open vacancies for this position, which is 1.5 times more than the vacancies of Data Scientist and Data Engineer combined. The shortage of good ML engineers in the world leads to the fact that the average salary for a market analyst in the United States reaches $141,000 per year, which is 25% higher than the average salary for Software Engineers who do not own machine learning technologies. We are for the balance of soft & hard skills and want to unleash the potential of the future ML engineer through the development of product thinking, so that the graduate can not only perfectly program and train deep neural networks, but also create truly successful AI products that solve real problems for end users.

Thanks to training through solving practical cases, support from mentors and internships with partners, graduates can count on the Middle position of a machine learning engineer (ML engineer), data engineer (Data engineer), developer of artificial intelligence systems (AI Developer), computer vision specialist ( CV Engineer) or Natural Language Processing (NLP Engineer).

What, in your opinion, can a university give to a technology company and vice versa?

At the moment, it is extremely difficult to adapt a graduate after online courses without real experience in production development. Therefore, we prepare undergraduates to meet the requirements of employers and on the basis of their technological stack, and the university will allow us to build the educational process.

Now talents do not want to study and work, but want to grow and develop, choosing a company where they will get quick experience, will be able to reveal their potential, therefore only the synergy of education and business can create conditions for training a holistic specialist in the modern VUCA world.

Similar Posts

Leave a Reply