Careers in Data Science: What you need to know

As educators, there’s no greater pleasure than in helping our students find their way in data science. A few months ago, we sent out a survey to all of our community members, asking them about their concerns in beginning a data science career.

And we received over 1,000 responses! On grouping them into common themes, the following questions emerged:

• Why do we need data science?
• What do I need to work in data science?
• How can I future-proof my job?
• When should I get started in data science?

In this first part in our series on careers in data science, we thought that we’d get a dialogue going by answering those pertinent questions right now. We would have loved the opportunity to ask these questions at the outset of our career. These and many more will be answered in the articles to follow, and in new and exciting products that will be released soon enough.

We’ll offer our finest hacks for getting your foot in the door, give you a ton of ideas as to where you can look for jobs, and show how you can start in data science with zero experience (it’s not as impossible as you may think!)

We love answering this question – data scientists are always looking for answers to the “whys” of business.

To reach a conclusion, we need to understand what the field actually is. For us, data science encompasses everything to do with data. That may sound broad when you consider how ubiquitous data is, but that’s simply what it is.

Once upon a time, data science could only be found as a component within other fields. Over the years, these fields developed their own silos of algorithms, methodologies, and communities that worked with data to answer business questions. But as data drives everything, it was only a matter of time that a discipline would emerge in its own right.

And emerge it has: The field has experienced serious growth, as we can see from this result of the search term “data science” on Google Trends:

Try it out for yourself. As with all great data science tools, you can filter your results (here, according to time, location, subject categories).

As the discipline can be applied to anywhere that has data, and as its principles add value, we can make a significant impact on the ways that businesses operate.

We need data science

• because data is everywhere
• because data adds value to businesses and people

As data science becomes more visible, many people have taken notice of the field, yet they remain mystified as to how they can access it. The good news is that the subject is so new that there is as yet no ‘standard’ way to get into the industry.

Hadelin de Ponteves and Kirill took different routes to becoming entrepreneurs in data science, so you may find you recognise yourself in one of our stories:

Hadelin began his data science journey in high school, when he developed a passion for mathematics. When it came to choosing a university degree, Hadelin naturally chose maths. Upon completing his undergraduate, Hadelin decided that he wasn’t yet finished with the subject and so he studied for a master’s degree in mathematics.

It was a logical move for Hadelin to then study at an engineering school in France. While there, he deepened his knowledge of a range of subjects, from maths to economics, biology, and even philosophy, before specialising in machine learning (ML) and applied mathematics.

To strengthen his portfolio, Hadelin also took the opportunity to intern at French premium television cable channel, Canal+, where he learned the ropes in finance and strategy. Ultimately, what Hadelin wanted was to find somewhere he could both continue with his education and apply his theories to business.

Shortly after his studies Hadelin was snapped up by Google, where he worked as a data scientist for their business intelligence unit. Working at a massive conglomerate like Google gave him immeasurable experience in deep learning. But his goal had always been to be an entrepreneur and so Hadelin decided that it was time to start making online courses for the next generation of data scientists.

The true key to his success? Hadelin swears by a large bar of chocolate, once a day! Eating mountains of chocolate isn’t going to turn you into the next Hadelin, but our takeaway from this is to do what makes you feel happy and energised. The journey to data science requires dedication and skill: don’t forget to be kind to yourself on the way.

Kirill’s story began the first time he sat down at a chessboard. He loved chess, particularly the logic and strategy behind the game. It always took ages for him to make a move because he was obsessed with analysing every eventuality possible for winning!

This early interest in logic was what drew him to physics at university. Although he enjoyed the subject, and it gave him some fundamentals that he could apply elsewhere, he didn’t continue with it after college because it would restrict him to one area. 

Additionally, Kirill had always shown interest in business-oriented subjects, more broad and versatile for whatever may come. It was for this reason that he undertook a master’s degree in accounting and finance.

Both of these subjects were fundamental pieces to build up the confidence that entrepreneurship was possible.

Even so, Kirill was cautious – he knew that to be successful he needed to see how other people did it. To acquire that knowledge, Kirill applied for a job at Deloitte’s accounting department and was lucky enough to get the job. On his first day, Kirill says, a Deloitte partner looked over his resume and said that he would be a great fit for their data science department – who was he to say no?

Kirill found working for Deloitte a fascinating study into how a professional services company operates, and was able to learn how he could excel in data science.

While Deloitte may be the golden goose for many people looking to get into the industry, don’t shoot only for the biggest names. Kirill had such a great experience there not because it is a well-known multinational but because they gave him the opportunity to apply his knowledge and boost his skills. This is what anyone new to the industry needs in order to excel – the chance to develop. Just select an industry and explore the many types of companies within it.

That’s what Kirill wanted from his next job: creating and building on the things he had learned. When he was headhunted by a pension fund, they had him develop their data science division from scratch, devise a way to help its organisation to work together on their data, and to deliver a different approach to customer experience.

He had a great time working there, but that got him itching to work for himself. Once he felt he was ready and had gathered enough experience, Kirill made the leap into entrepreneurship and started to create Udemy courses on data science.

There is no single route to data science, so the world is your oyster. For us, the key is to ensure that you have a balance of education and experience in order to become knowledgeable and confident in the industry, but it’s up to you as to where and how you acquire them.

Find knowledge in others experiences, read up as much as you possibly can, engage with others that have been where you want to be and have fun while trying to achieve your own Data Science goals!

As data grows exponentially, it stands to reason that we will need more data scientists to produce the tools that can handle it. This might make us believe that a career in data science is absolutely safe from automation.

This, however, is not the whole story – the better we get at producing tools to cope with data, the more automated we make our own field. There are robots that can handle the entire Data Science Process, without any need for human interaction.

So, will data scientists get replaced by their own algorithms?

Hadelin and Kirill have discussed this exact matter with Jeremy Achin, CEO of DataRobot. Achin and his team have created a self-serve ML platform where anyone can upload their data for a machine to process. These robots are hardly rudimentary. They allow you to get insights not only from your inputted data but also for a specific problem in need of a solution. DataRobot is one of many companies automating data analysis.

We know the way that the field is moving, and we don’t want to sell you any false dreams that, if you become a data scientist, your job will be 100% secure. The way to keep ahead of the game is to understand that while a data robot may indeed push out the number crunchers, data science is also reliant on people dreaming up new ways to capture, store and process information.

A machine might be able to do things more quickly, and even to identify the optimal solution for a task, but it cannot (yet) devise entirely new approaches to a data science project.

Think about it – is it harder to follow the rules to get results, or to work out entirely new pathways to achieving them?

Getting into the field right now is the best time to upskill yourself to a level where you aren’t doing the ‘at risk’ tasks, protecting you against the threat of automation. The best advice we can give is therefore to focus on strengthening your creativity.

For more insights into future-proofing your career, there is only one real advice to follow: keep yourself curious, keep yourself hungry and learn as much as you possibly can. Even though there is a data tidal wave moving this way, we don’t want to alarm you, with it comes hundreds of new and exciting opportunities. You only need to be prepared and the best way is to get started right away!

Robots will become smarter, and data science isn’t entirely immune to automation. But we also need data scientists to understand the mechanics of these robots, bots and algorithms: how do their functions work? how they can be managed? how can we get actionable outputs for companies? We also need thinkers – people who can take on the creative tasks that machines are less capable of handling. Get into the game early, and you’ll be streaks ahead when automation starts to happen.

Considering the field is in its relative infancy, some people wonder whether it might be better to wait until it has grown before getting their feet wet. If you’ve read the previous point, you should know our answer to this question: there’s no time like the present.

Companies are starting now to introduce data science teams to their organisation – even those whose core output has nothing to do with science! We live in a highly competitive world. Savvy companies understand that, to stay in the game, they need a data science team to gain insights into customers and business operations that give them a competitive edge, ensuring they aren’t driven out of the market.

This means that when one, just one, company starts to us data science, its competitors will have no choice but to follow suit.

As most companies are now creating data science teams, more opportunities will emerge. Quant Crunch, a Burning Glass Technologies report commissioned by data science leaders IBM notes that:

[image 7]

If we consider that the population of the US is close to 400 million, this would effectively mean that well over half a percent of the US population will be in data science by 2020! Even now, data science jobs typically remain open for 45 days, which is “five days longer than the market average” (Burning Glass Technologies, 2017). This means that it’s 10% harder for companies to fill these roles, which tells us that the personnel are currently not there.

So, what are you waiting for? Get into the field now however you can, whether it’s in education or practice, and ensure you’re a part of the community of the future.

There are so many ways to get started, and you can do so today: Why not take a course, grab a couple of good books, become an intern, practice with real-world datasets, or help out in a citizen science project? Even if you’re ‘only’ educating yourself at this stage, you’re still keeping yourself in the game and in the community. And if you’re just biding your time, don’t. Companies are looking for you, and they’re willing to pay you handsomely.

How do you feel about the future of data science as a discipline? Where do you see the most opportunity for data scientists? If you’re already a practitioner, do you have a different view of any of these questions? Let us know, and let’s get the conversation going!