According to research by HeadHunter and Mail.ru, the demand for professionals in Data Science exceeds supply, but even so, young specialists often struggle to find work. We explain what graduates of these courses are missing and where to study for those planning a significant career in Data Science.
"They come in thinking they will earn 500k a second because they know the names of frameworks and how to launch a model in two lines of code."
Emil Magerramov leads a group of computational chemistry services at biocad and encounters candidates in interviews who lack a systematic understanding of the profession. They finish courses, come in with strong Python and SQL skills, can set up Hadoop or Spark in two seconds, and execute tasks according to a clear specification. However, when it comes to flexibility, they fall short. Yet it’s exactly the flexibility that employers expect from their Data Science specialists.
What is happening in the Data Science market
The competencies of young specialists reflect the state of the labor market. Here, demand significantly exceeds supply, so desperate employers are often willing to hire completely inexperienced specialists and train them. This approach works, but only if there is already an experienced team lead who can take on the task of mentoring the junior.
According to research by HeadHunter and Mail.ru, data analysis specialists are among the most sought after in the market:
- In 2019, the number of job vacancies in data analysis increased by 9.6 times, and in machine learning – by 7.2 times compared to 2015.
- Compared to 2018, the number of job vacancies for data analysis specialists increased by 1.4 times, and for machine learning specialists – by 1.3 times.
- 38% of job openings are in IT companies, 29% in financial sector companies, and 9% in the business services sector.
The situation is being fueled by numerous online schools that train those very juniors. Typically, the training lasts between three to six months, during which students manage to acquire a basic understanding of the main tools: Python, SQL, data analysis, Git, and Linux. The result is a classic junior: capable of solving a specific task, but still unable to comprehend problems and formulate tasks independently. However, the high demand for specialists and the hype surrounding the profession often lead to high ambitions and salary expectations.
Unfortunately, a Data Science interview nowadays usually looks like this: the candidate recounts that they’ve tried using a couple of libraries, but when asked how the algorithms actually work, they can’t provide an answer, yet they request 200, 300, or 400 thousand rubles per month take-home pay.
Due to the overwhelming number of promotional slogans like "anyone can become a data analyst," "master machine learning in three months and start earning a lot of money," and the desire for quick gains, our field has seen an influx of superficial candidates who are completely unprepared systems-wise.
Victor Kantor
Chief Data Scientist at MTS
What employers are looking for
Any employer would prefer their juniors to operate without constant supervision and be able to grow under the guidance of a team lead. To achieve this, a novice must immediately possess the necessary tools to tackle current tasks and have a sufficient theoretical foundation to gradually propose their own solutions and tackle more complex problems.
The newcomers to the market are fairly well-equipped with tools. Short-term courses allow them to quickly master them and start working.
According to research by HeadHunter and Mail.ru, the most sought-after skill is proficiency in Python. It is mentioned in 45% of job postings for data analysis specialists and in 51% of job postings in the field of machine learning.
Employers also want data analysis specialists to know SQL (23%), be skilled in data mining (19%), understand mathematical statistics (11%), and be able to work with big data (10%).
Employers looking for machine learning specialists expect candidates to be proficient in Python, along with C++ (18%), SQL (15%), machine learning algorithms (13%), and Linux (11%).
However, while juniors are well-equipped with tools, their managers encounter another issue. Most graduates of courses lack a deep understanding of the profession, making it difficult for newcomers to progress.
I am currently looking for machine learning specialists to join my team. However, I see that often candidates have mastered individual Data Science tools, but they lack a solid understanding of the theoretical foundations necessary to create new solutions.
Emil Magerramov
Group Leader of Computational Chemistry Services, Biocad
The structure and duration of the courses do not allow for the necessary level of depth. Graduates often lack the very soft skills that are usually overlooked when reading job postings. Honestly, how many of us would say we lack systematic thinking or a desire to grow? However, when it comes to a Data Science specialist, it’s a deeper issue. To develop, a strong focus on theory and science is needed, which is only possible through extended education, such as at a university.
Much depends on the individual: if a three-month intensive from strong instructors with experience as team leads in top companies is taken by a student with a solid background in mathematics and programming, who absorbs all the course materials like a sponge, as we used to say in school, then there are no problems with that employee later on. But for 90-95% of people, in order to truly master something, they need to learn ten times more and do it systematically for several years. This makes master's programs in data analysis an excellent option for gaining a solid foundation of knowledge, which ensures that one won't feel embarrassed in interviews and makes the job significantly easier.
Victor Kantor
Chief Data Scientist at MTS
Where to study to find a job in Data Science?
There are many good courses available in the Data Science market, and acquiring initial education is not a problem. However, it's important to understand the focus of this education. If a candidate already has a strong technical background, then intensive courses are just what they need. A person will master the tools, come into the position, and quickly adapt because they already think like a mathematician, recognize problems, and formulate tasks. If such a background is lacking, they'll become a good executor after the course but with limited growth opportunities.
If you're facing a short-term goal of changing professions or looking for a job in this field, then you should consider systematic courses that are short and quickly provide a minimum set of technical skills, enabling you to qualify for an entry-level position in this area.
Ivan Yamshchikov
Academic Director of the Online Master's Program in Data Science
The problem with courses is that they provide a quick but minimal boost. A person literally jumps into the profession and quickly hits a ceiling. To enter the profession for the long term, a solid foundation needs to be laid right away through a more extended program, such as a master's degree.
Higher education is suitable when you understand that this field interests you in the long run. You're not rushing to get a job as quickly as possible. You also don't want to face a career ceiling, nor do you want to encounter the problem of lacking knowledge, skills, or a comprehensive understanding of the ecosystem through which innovative products are developed. This requires higher education that not only provides the essential set of technical skills but also restructures your thinking and helps shape your vision of a long-term career.
Ivan Yamshchikov
Academic Director of the Online Master's Program in Data Science
The absence of a career ceiling is the main advantage of a master's program. Over two years, specialists acquire a solid theoretical foundation. Here's what the first semester of the Data Science program at NITU 'MISIS' looks like:
- Introduction to Data Science. 2 weeks.
- Fundamentals of Data Analysis. Data Processing. 2 weeks.
- Machine Learning. Data Preprocessing. 2 weeks.
- EDA. Exploratory Data Analysis. 3 weeks.
- Key Algorithms in Machine Learning. Part 1 + Part 2 (6 weeks)
At the same time, it is possible to gain practical experience at work. Nothing prevents a student from applying for a junior position as soon as they have mastered the necessary tools. However, unlike a course graduate, a master's student does not stop their education here but continues to deepen their professional knowledge. In the future, this allows for unrestricted development in Data Science.
The website of the National University of Science and Technology 'MISIS' hosts for those who want to work in Data Science. Representatives from NITU 'MISIS', SkillFactory, HeadHunter, Facebook, Mail.ru Group, and Yandex discuss the most important topics:
- 'How to find your place in Data Science?',
- 'Can you become a data scientist from scratch?',
- 'Will the demand for data scientists remain over the next 2-5 years?',
- 'What tasks do data science specialists work on?',
- 'How to build a career in Data Science?'
Online training, state-accredited diploma. are accepted until August 10th.
Source: habr.com
