As more and more companies implement machine learning and AI systems, we’re seeing a growing need for data scientists and data science teams in a variety of fields. This can be seen in a whole manner of industries, from ecommerce and medicine to transport and warehouse robotics. This growing need for data scientists also comes with a need for a robust education system for developing data science skills.
In order to get a clearer understanding of what companies are looking for in data science positions and how education is addressing those needs, I got in touch with Le Wagon Tokyo, who recently announced a Data Science Bootcamp covering a spectrum of data science tasks from coding and data collection to building deep learning systems.
In this discussion with co-founder Sylvain Pierre and data science instructor Trouni Tiet, we cover the current state of data science toolkits, how the Le Wagon course prepares students for future careers, and current trends in the field of data science.
What is Le Wagon?
According to Sylvain, Le Wagon is a product-oriented coding bootcamp with a hands-on approach, where students spend 90 to 95% of their time building products. “Our objective is not to just churn out engineers,” he says. “Though we do train a lot of people who become engineers, really we teach people how to build products.”
Founded in France in 2013, Le Wagon began under the motto of “bringing technical skills to create entrepreneurs”. Their web development course is taught in over 20 countries and boasts over 8,000 graduates across 70 nationalities. Their recently announced data science course is scheduled to begin in Japan on October 12th, 2020. It aims to give graduates the skills to collect, store, clean, transform, and predict data in production environments.
The Le Wagon Data Science Course
The Le Wagon Data Science bootcamp was developed over the course of a year. The course was designed in collaboration with tech company data science professionals and researchers in AI.
“There was definitely a demand for the course, because you can see that companies are really struggling to find engineers,” says Sylvain. “But I would say that the real driver of the bootcamp was the toolkits. Look at the job descriptions for data related positions; the toolkits they ask for are almost always the same.”
With this in mind, Sylvain pointed me to how Le Wagon’s data science course covers the industry standard languages, tools, and libraries. This ranges from programming languages like Python and SQL to the machine learning and deep learning libraries scikit-learn, keras, and TensorFlow.
“I think the most important thing is to keep a very practical approach,” he says. “It’s not just about understanding the theory, it’s about really knowing how to bring actionable results for companies. The bootcamp is never going to be just an explanation of a topic; the idea is always let’s talk about this stuff, do it together, and make sure you can do it by yourself.”
Trouni adds, “You can think of [the bootcamp] as two months in the life of a coder or a data scientist. More than teaching to make you learn, we’re teaching you then making you apply those teachings. Putting theory into practice is crucial.”
The Importance of a Hands-on Approach for Data Science Careers
Le Wagon puts a strong emphasis on building and creating products as part of the bootcamp experience. According to Trouni, building products is key to making up for the experience gap that many face when looking to start a data science career.
“Companies often ask for two or three years of experience. But many students don’t have this after completing a bootcamp,” he says. “And this much is true: it’s hard to compete with someone who has studied and has work experience on their resume. But that doesn’t mean you can’t do what they do. You can still catch up.”
“So I think there’s a huge opportunity for people breaking into the field to sell themselves on the strength of their portfolio. You may not have the two years of experience, but that might not matter; your portfolio is a chance to show you can do the work; you can build a product.”
This idea of building a portfolio you can take to the career market after completing the bootcamp is important to the Le Wagon approach. It also fits with their research into the data science market and its needs.
The State of the Data Science Market
When I asked about the data science market in Japan, Sylvain pointed out that many small and medium-sized startups are well aware of how important it is to leverage data, but don’t have the budget to hire a comprehensive data science team.
“These sorts of companies are looking for multi-talented people who can play several roles at the same time. It’s an important aspect to keep in mind,” Sylvain says. “I also think we’re seeing a trend towards companies having at least one person who understands data science. Many companies still don’t have someone to fill that role, so what we’re seeing is a need for data scientists who are able to pull actionable business insights from data.”
This fits with Le Wagon’s research, which has also shown that more than 50% of their hiring partners are looking for data analysts, data scientists, and data engineers. And because of the broad needs across the landscape, being a jack of all trades can be a strong starting approach before developing a specific field of expertise.
A Look at Data Science Trends
Trouni was quick to point out that the rapid progress in machine learning and AI makes pinpointing trends very difficult. “It’s really subjective, because it depends on the topics you’re interested in,” he says. “But I’ve noticed that everything is moving. It’s not just one specific industry, but all of them are moving and growing; even fields like agriculture, which was once considered a niche field.”
As for his own areas of interest, Trouni pointed to developments in computer vision. He mentioned the creation of 3D environments from photographs and the use of AI to create lifelike landscapes from basic colors and shapes. In this he sees the development of AI as a collaborative tool to automate tasks and make them simpler.
“I think [this kind of automation] can be applied everywhere,” he says. “So it’s more than just the trends that interest me personally. What we’re seeing is countless other trends, and they’re all moving as a front.”
When I put the same question to Sylvain, he pointed to the accessibility of the field. “I think we’ve started seeing that you can be a data specialist, engineer or analyst, and not necessarily be a researcher,” he says. “What I mean is, [data science] is becoming more accessible to both individuals and companies. You don’t have to be a data researcher to get value from data science today. The person in charge of data in your company doesn’t necessarily need to know how to build [machine learning] models. Yes, they have to know how it works, but with the tools we have now, that person mainly has to be very good at analyzing a company’s data and bringing their business mindset to it.”
The Le Wagon Data Science Bootcamp begins on October 12th, 2020. You can learn more about the Le Wagon Bootcamp at their official page, which details available locations, the bootcamp structure, and the syllabus. For more articles about data science projects and machine learning techniques, be sure to check our resources page for more.