Advantages of the program
Machine learning engineers are developers who develop and train artificial intelligence. Voice assistants like Siri and Alexa and personalized social media feeds – these are the examples of their work. Such specialists are currently in high demand on the global labor market. In Russia, according to habr.com and jooble.ru, their salaries amount to approximately 1.8 million rubles per year (some 150 thousand rubles per month); in the US – $141 thousand per year.
The program was launched as part of the AI Talent Hub (link in Russian) project, a winner of the open call held by ITMO within the framework of its 2030 Development Strategy. The project’s goal is to achieve a new level of specialist training in the field of AI for middle specialists through distance learning and project work. By 2024, the AI Talent Hub team plans to ensure employment of 60% of graduates in middle+ positions in AI, as well as provide 10 programs of varying formats (online Master’s, online courses, and microdegrees).
The new Master’s program is designed for students aspiring to become machine learning engineers. Students can expect a whole new learning format based on top methodologies from engineering schools, practices used to support interns at IT companies, and interactive online learning technologies.
A major part of the program will be implemented online, allowing students to learn from anywhere in the world. Studies will take place in the format of a distributed project office and under the supervision of experienced mentors.
According to Dmitry Botov, the head of AI Talent Hub, the idea for the program came from his own experience with teaching machine learning in the flipped classroom format and hosting corporate practice-focused courses.
“I spent over 10 years at various IT companies, universities, and online schools, analyzing the ways in which students, recent graduates, and established specialists deal with new approaches and technologies. Allowing experienced mentors from the industry to collaborate with students on projects produces a positive synergistic effect. Flipped online learning helps, too: students master theory at home while classes are almost entirely interactive – we break down cases, work on projects, and review solutions,” he explains.
One of the program’s industrial partners is Napoleon IT, a company that specializes in the development of cloud computing-based high-load systems.
“We’ve been collaborating with ITMO for a long time; we can always expect a completely unconventional, flexible, and modern approach from the university. Only this kind of collaboration can produce breakthrough projects like AI Talent Hub. We don’t just want to provide practical training; we want to provide working conditions in which the students can grow as professionals. This is, after all, what happens when a fledgling specialist joins an experienced team. We hope to turn this into a world-class educational initiative,” comments Pavel Podkorytov, a co-founder of Napoleon IT and CEO of TalentService.com.
Students who have completed the program will be able to build their career as machine learning, data, computer vision, or NLP engineers, as well as AI developers. Moreover, their qualification will correspond to the middle specialist level of leading companies – after all, the program is based on the analysis of the labor market and its demands.
The program pays great attention to striking a balance between hard and soft skills, as well as developing the graduates’ product thinking: graduates must be able to not only program services and train neural networks, but also create AI products that would solve real issues.
“I won’t speak for the whole market sector, but our ML-focused company, which employs over 100 staff members, is looking for specialists who have practical experience with deploying models and applying ML in production. We always hire those who are capable of creating scalable ML services. But now is the first time that we can participate in their training, too: familiarize them with our tech stack and cases and involve them in our business processes. It’s important for us that these talented kids are able to start working on middle engineer-level tasks while they’re still at university,” says Pavel Podkorytov.
As part of various project teams, the students will go through all stages of AI development: from collecting and labeling data and mapping its distribution routes to training deep neural networks, developing backend services, and integrating them into UI via APIs. The program requires students to complete at least six project modules. In order to complete each one, they will need to present a functioning software service that meets all requirements set by the client.
Other learning elements change, too: lecturers become project module authors, lectures turn into introductions to each week of a module, lab classes are replaced with expert guidance, and exam questions are swapped for project defenses and interviews focused on assessing a student’s soft and hard skills.
Upon completion of their first year, students will have become familiar with classic methods of machine learning, computer vision technologies, and natural language processing. The first year will also culminate in a three-month internship in a project team at one of the program’s industrial partners. In their second year, students will work on a graduation project in one of several formats: they may choose to develop AI solutions with a partner company’s project team, conduct an applied study using data science methods, join an open-source project, or launch a startup at an AI accelerator.
“We’re talking to a number of Russian and international IT companies and research centers that are looking for ML engineers and AI developers. Artificial intelligence is a key component of their products and technological processes. We’re also talking to our partners among the leading companies in computer vision, conversational AI, and predictive analytics about launching project modules in their fields and organizing internships. We also see great value in collaborating with regional IT companies that have their own centers of expertise and industry-specific projects,” comments Dmitry Botov.
How to apply
The program is suitable for practicing IT specialists as well as those thinking of transitioning into the field of machine learning. Applicants must be familiar with mathematical analysis, linear algebra, probability theory, statistics and basic ML and neural network algorithms, as well as know how to program in Python.
To apply, prospective students can take a conventional interview/exam, participate in the report contest at the Congress of Young Scientists, or take part in ITMO University’s Portfolio Contest and the Mega Contest.