We recently caught up with Dmitry Repin, Founder at New Professions Lab. We will be chatting about their approach to Data Science education and also explore the different offerings at Newprolab.
Hi Dmitry, Thanks again for taking the time to join us for this interview. Let’s kick off with a few introductory questions
Q : What’s your 1 minute bio / introduction?
A : Dmitry Repin — an academic turned entrepreneur. My past experience includes running Digital October in Moscow as CEO from day one and 3 people in a conference room to an organization with 40+ full-time employees with $7M annual revenue. I was involved as a Co-founder in launching and running business incubators (HSE Inc), entrepreneurial competitions (BIT Competition), ecosystem centers (Digital October), educational non-profit projects (Innovation Design Lab, Knowledge Stream), educational for-profit ventures (New Professions Lab, Start-in-Garage), international conferences (TechCrunch Moscow, Landing Page, The Art of Going Global, LP Unlimited).
Media involvements include producing niche tv-shows (Chain Reaction, October Evolution), producing and co-hosting radio-shows (Dobry Venture, Let Them Teach Me First). My academic work involved doing research and teaching in Boston (MIT Sloan), Gotenborg University, Moscow (Higher School of Economics, Moscow School of Management SKOLKOVO, New Economic School).
Since 2007, we have been running eXtrepreneurship, a program about being an entrepreneur while sailing for one week in a 45-50ft sailboat in various remote locations on the planet.
Q : Can you tell us a little about your background and role at Newprolab?
A : I’m the sole remaining founder at Newprolab — at times there were two other partners but at different points, in time they went on to pursue other endeavors. My current role is sort of a mastermind, I’m involved in strategic decisions and international business development. I guess my entrepreneurial background helps to cut the crap out of the content and get to the point, while academic background helps to cut the bullshit out of the content and link it to the rest of the knowledge universe.
Q : How did you get into Data Science Education / Training?
A : It turned out to be a niche that was mostly underserved when Newprolab started and it has remained an area where our competence is accepted worldwide.
Q : What is your definition of a Data Scientist?
A : In a narrow sense meaning, a data scientist is a person who is responsible for extracting knowledge from preprocessed data. Knowledge might be a report full of different insights and prescriptions or it might be a machine learning model (a representation of a small part of our reality).
Q : What’s the 1 minute bio / introduction on Newprolab?
A : Newprolab offers in-depth practical education for professional audience who aspire to become data professionals.

Q : How did the idea for Newprolab come about and what do you hope to achieve with it?
A : The idea came as something that was needed for the Russian technology audience to be more competitive in the global market. Newprolab’s goal is to be a leader in high-end, in-depth, data-related education worldwide.
In general there are two profiles. The first one is with a technical background – they have experience in software engineering, at least during their university years. The second one is more grounded in math and statistics. Our programs suit both of them.
Q : How do you screen and select fellows for your program?
A : We have some entry quizzes to measure how well a fellow is prepared for the program but that serves only an informational purpose. If a person got bad results but still thinks that they can go through the rigors of the program realizing it will be challenging, we don’t prevent them. Actually, we have examples where such fellows became top performers in the cohort.
Q : Can you describe the typical background (academic/professional) you look for in your fellows?
A : In general there are two profiles. The first one is with a technical background – they have experience in software engineering, at least during their university years. The second one is more grounded in math and statistics. Our programs suit both of them. However, there are different topics in the programs that are hard for each group – for software engineers it’s hard to understand math and statistics, for analysts – to write code.
Q : What type of skills or traits do you look for in a prospective fellow?
A : Grit. Our programs are not easy and require a lot of autonomous work. A prospective fellow should have grit to overcome these difficulties as that’s the only way to acquire skills. We created a system that helps and motivates people to grow.
Q : How many cohorts have you gone through?
A : It depends on the program. We have several open-enrollment programs related to data: Big Data Specialist (8 cohorts), Deep Learning (4 cohorts), Big Data for Executives (3 cohorts), Data Engineer (2 cohorts) – with total number of about 350 participants. Besides, we have run several corporate programs in Russia and Luxembourg with 260+ people.
We never guarantee job placements after attending our programs as it’s a rather hard and complicated process with many factors including their professional background and soft skills. We guarantee that we’ll do our part of the job – help them acquire relevant skills and knowledge of up-to-date instruments and technologies
Q : Do you run multiple cohorts at the same time?
A : Yes, we do run 2-3 different programs at the same time and even in different countries Russia and Luxembourg). Now we have Big Data Specialist and Data Engineer programs in Moscow. In the fall 2018 we plan to launch an open-enrollment Data Engineer program in Germany, so we’ll be operating in several countries.
Q : What is your typical cohort size?
A : Well, it depends on the program. Big Data Specialist and Data Engineer usually have 25 – 40 fellows. Deep Learning and Big Data for Executives have about 10 – 20 fellows.
Q : Can you share what your placement numbers look like for your most recent

cohort?
A : It’s an interesting question. Not all of our students plan to transition into a new job. For example, only 15% of all students in Data Engineer program declared they were thinking about changing companies but they weren’t quite sure. So, most of them just want to grow in their current company. Besides, we have a significant proportion of fellows whose participation is covered by their employers. On average 1 to 3 students in a cohort plan to transition to a new job within three months post-graduation. Sometimes their fellow classmates invite them to join their teams, our alumni are also eager to enrich their teams with the people graduated from our programs.
Q : Can you also share your historical placement rate (within three / six months of graduation of each cohort)?
N/A
In our opinion, it’s crucial that a professional in this field understand and learn new things or tools on their own.
Q : What percentage of your fellows eventually get Data Scientist vs Data Analyst vs Other technical jobs?
A : We have 3 career paths after our main program (Big Data Specialist): Data Scientist, Data Engineer, Data Manager. The proportion is : 67%, 20%, 13% respectively. Data Analysts usually come to our program to acquire some new skills and knowledge to enable them transition to a new role.
Q : Can you share with us where some of your graduates work or will work?
A : Most of our alumni are Russia-based. However, we have some that moved to other countries like Germany, Ireland, Netherlands and USA. We also had online participants from CIS countries, Latvia, Estonia, Israel, Great Britain, Poland. You probably won’t recognize many brands, but I’ll try to recall the international ones: Airbnb, Booking.com, Zalando, XING, VISA, Accenture, Microsoft, NVIDIA, Yandex, Mail.ru Group, Sberbank, MegaFon, Raiffeisenbank, Alfa-bank.
Q : Can you share with us what industries your graduates work in?

A: Top-3: banks, telecoms, e-commerce. They are innovators or early adopters of new technologies in Russia. Plus, of course, several huge tech companies like Yandex and Mail.ru Group.
Q : How do you prepare your fellows to be very competitive for Data Science jobs?
A : In our opinion, it’s crucial that a professional in this field understand and learn new things or tools on their own. We simulate real business environment instead of creating a greenhouse for the fellows. Imagine you are an employee in this field who was given a task to do something that no one had ever done in your company before. You have to deal with this huge uncertainty and ambiguity and then find a way to solve the task. We implement this in our practical assignments (laboratory tasks as we call them) – some parts of them are thoroughly described, some are white spots and you have to explore and actually google the next steps. In our experience, such professionals are in high demand and may have a higher salary.
In my opinion, there are two types of situation: if you hire a trained specialist, you need to assess their background and ability to solve your tasks. If you hire a junior-level professional, then you need to assess their ability to grow. It includes grit, determination, general ability to learn something new and good soft skills
Q : Do you have a hiring day and what percentage of students are typically placed from a company they meet at hiring day?
A : As I’ve mentioned before, only 15% of participants in a cohort want to find a new job. We never guarantee job placements after attending our programs as it’s a rather hard and complicated process with many factors including their professional background and soft skills. We guarantee that we’ll do our part of the job – help them acquire relevant skills and knowledge of up-to-date instruments and technologies. At the same time we understand we may help those who want to get a new job (fellows of current programs and our alumni), so earlier this year we have partnered with a top recruiting company in the IT field in Russia – Spice IT Recruitment. Twice a year we organize a career day for our alumni and current fellows. At these sessions, representatives of Spice IT Recruitment share information on current trends, advise on how to write CVs and prepare for an interview, and will individually work with participants answering their questions. The agency receives CVs directly from our alumni and helps them find their dream job.
Q : For organizations looking to hire Data Scientists what should they look for..Ivy Degrees, PhDs, Extensive experience, Quantitative Background, Technical chops, grit or determination?
A : Of course, it depends. Currently, the demand is higher than the supply. To look for degrees and PhDs means to reduce your opportunities to hire someone helpful for your business. In my opinion, there are two types of situation: if you hire a trained specialist, you need to assess their background and ability to solve your tasks. If you hire a junior-level professional, then you need to assess their ability to grow. It includes grit, determination, general ability to learn something new and good soft skills.
Q : Can you give a short summary of a typical day in the life/week in the life for your fellows?
A : Ooooh, we better ask them (participants of Big Data Specialist and Data Engineer programs). First of all, they all work full-time and have families. How do they manage that plus the program? They’re supermen and superwomen. They have 3-hour evening classes 3 times a week and laboratory task for the week with a strict deadline. And as you remember they have to find answers within the lab on their own (but they share information in the group chat and help each other) and it takes time as well. But they all manage and according to their feedback, it is such a relief after the program to realize how much free time you have!

Q : How do you improve your process from cohort to cohort at Newprolab?
A : I guess, not surprisingly we get feedback. In the short-term, we ask the fellows after each class how they liked it. In mid-term, we ask them about the likelihood to recommend the program after graduation and also what they liked and didn’t like. In long-term, we measure the retention rate (how many of our alumni go to another program) and the proportion of those who came by a recommendation.
Q : What skills and tools do you think should be emphasized more in Data Science education?
A : I think it’s quite important to understand the business. How to choose the right metrics, for example. Changing a metric may increase your revenue by several percent. Or how to evaluate the effect of your model or proposition. Sometimes you can come up with an idea that building or implementing a model is not even worth doing. It’s important. It’s not about the algorithms, but it’s about the reality.
Q : Given the very fast progression in the field, what skills do you think will be most important for Data Scientists in the next few years?
A : I think it will be important to not only train your models Kaggle style but to also ensure you can make them production-ready. For instance, in our program, we have a competition where you have to meet some SLA (40 milliseconds) for your model. You can compete with your scores like AUC or other metrics only once your model has met the SLA requirement.
Q : What makes Newprolab unique and how do you differentiate Newprolab from the other offerings out there?
A : We have a broad range of data-related programs. Other bootcamps usually have fewer programs and mostly focus only on the data science part. We are trying to cover the whole universe of skills and roles in the data field. We also have a strong hands-on approach. We are making our fellows’ lives easier by preparing them for the real world.
Q : How do you help your fellows deal with burn-out?

A : They don’t burn-out if the programs are designed properly. Your first practical assignments shouldn’t be super hard. It’s important to give the fellows some positive feedback at the start: you know little but you already can do some things that you should be proud of. And then you can gradually increase the complexity level. Also, we have a strict schedule that also helps to keep them in and prevent from postponing the tasks.
Q : Are fellows ever asked to leave or are kicked out mid-way through the program or at anytime during the program?
A : No, never. However, we had cases when a fellow came to us to say he couldn’t cope with the program as it’s too hard and they don’t have enough time.
Q : Can you give us a sample of the tools, languages, and techniques your fellows are exposed to during the program?
A : Our main language is Python. In Big Data Specialist program fellows deal with the Hadoop ecosystem tools (HDFS, HBase, Hive), Spark from the user side, Python libraries like scikit-learn, matplotlib, pandas, gensim, NLTK, etc. In the Data Engineer program they work with Kafka, Elasticsearch, Logstash, Kibana, Spark, ClickHouse, Airflow, Superset, Grafana and Prometheus. In Deep Learning program, they use Keras and a bit of Caffe.
Q : What do you feel is broken with Data Science education in general and do you have any suggestions on how it can be improved?
A : Yes, there are some things that are broken, but I don’t think that there is something special in data science education. Usually, people design programs about several topics that they are trying to cover. In my opinion, we should focus on real tasks from business implications and then inject those topics and practical assignments to drive these tasks, but not vice versa.
Q : Do you have a structured alumni program?
A : Yes, more or less. We realized that our alumni network is really important for our business. For now, we have a referral program. You can invite your friends or colleagues to our program and get some points that you can use to pay for other programs. We have regular monthly informal meetings at a pub. Sometimes, we have special offers for conferences and hiring days.

Q : Do you support your fellows after they’re done with the program?
A : Probably not in a structured way, but yes. We have a Facebook group where everyone can ask questions and expect that someone will answer. Our instructors are also in this group.
Q : Do Newprolab alumni stay involved with the program and help make introductions/referrals for new fellows?
A : Yes, in the opening of the main programs we have a tradition where our alumni give some advice and lifehacks about the program. Some of them even get back to us and become instructors while keeping a full-time job.
Q : What problems in Data Science / Data Science Education keep you up at night?
A : Well, usually I have a good sleep at night. But probably for me as a founder of a bootcamp in the field, it is important to be in touch with the current trends in the market and especially for the new roles. For example, I’m thinking right now about the differences between Data Scientist and ML Engineer roles.
Q : Have you faced any major challenges in running Newprolab?
A : Not really.
Q : What markets/verticals are you currently focused on?
A : I can’t say that we have some strong focus on any vertical. We have good operational experience in Russia. We are currently expanding to Europe and I think we can bring something useful to this market too.
Q : How do you feel the job market differs across different industries / Europe?

A : Companies in different industries have different goals. Thus, they require people with different skills. For example, in manufacturing one of the most popular tasks is anomaly detection, in e-commerce — recommendations, in banking — scoring and anti-fraud. Those tasks are solved by different algorithms and require slightly different mindsets in data scientists. Another important factor is the size of the company. If it’s big, you’ll probably see many different roles and specializations. In a small company, you’ll find generalists who can both analyze data and write production code.
Q : Who are the Data Scientists that inspire you?
A : It’s a hard question. Probably, they are the people who are at the intersection with other domain knowledge. As an entrepreneur, I’m inspired by the data scientists who are also entrepreneurs: Andrew Ng, Daphne Koller, and Sebastian Thrun. I think it’s not a coincidence that they are into education too as I have my own biases.
Q: Any parting words for prospective Data Science students or Data Scientists that are just starting their careers?
A : Think about your strengths. You can probably be indispensable not as a Data Scientist. There is nothing bad or wrong with that. The Data Science profession is just more advertised in the media. If you are good at writing code or reading documentation, you can be really good as a data engineer. If you have good communication skills and business knowledge, you can be a good business analyst. There are a wide variety of roles out there and you can become proficient at any of them. Don’t limit yourself to data scientist role.
Thanks, Dmitry for taking the time to share some of your insights and discuss New Professions Lab with us.
To find out more about New Professions Lab you can either reach out to Dmitry Repin, engage with Newprolab on facebook or reach out to their former students or Instructors. Newprolab is currently expanding their programs across Europe.
Also, please stay tuned for the other Data Science Bootcamps Founder Interviews we have in the pipeline at Data Science Bootcamp Founders Interview Series