How to Become a Data Scientist

What is she doing?

Data science is one of the hottest fields right now and shows no signs of slowing down. By 2018, McKinsey estimates that American businesses will need over 490,000 data scientists, but there will be fewer than 200,000 qualified applicants to fill those positions.

Where do you apply, right?

Not so fast. It takes a certain set of skills and education to be a data scientist, said Chris Neimeth the COO of NYC Data Science Academy, which helps students and companies learn data science skills and manage their data science needs. “A data scientist is a storyteller, an engineer, and they’re a business analyst; they need to understand what the problem is that the business is trying to solve.” You also should like math. A lot.

“Generally people who don’t have a comfort with numbers and the ability to write code are probably not going to be successful,” said Neimeth. Most of the people who take the NYC Data Science Academy bootcamp come from a STEM background. 80% have masters degrees and 30% have PHDs, said Neimeth.

Still interested?

Soon almost every imaginable type of business will be looking for a data scientist. These jobs have become more plentiful thanks to a confluence of factors. Businesses’ data is increasingly becoming digitized, computers are faster and cheaper with higher computing power, and the internet of things is bringing new sets of data for data scientists to sift through.

Data science is now being used in finance, healthcare, retail, marketing, government and other industries.

So what’s the job like?

At quantitative hedge fund Two Sigma, data scientist Ben Wellington said no two days are the same. “One day I might be out looking for various data sets about traffic. The next I might be cleaning data about ice cream consumption! Another I might be testing out a hypothesis I have about weather. It’s really diverse work.”

He landed his role as quantitative analyst after studying computer science and math and finishing a Ph.D. program in computer science.

“I think the most important quality to display when looking for a data science role is passion for data,” said Wellington. “We love people who do independent data projects just because they are curious to learn more about the world.” But don’t forget about the science. “Not enough people make sure that their published work is reproducible, and that’s a shame. It is one of the core tenants of the scientific method.”

Although the term data science has exploded recently, the science has been around for a while, said Wellington. “What has changed is the amount of demand, and the amount of freely available, high-quality software to do analysis.” When Wellington started his job 10 years ago C++, JAVA, and Perl were the programming languages of choice. Today it’s Python and R.

In addition to that software ability and passion for business, math, and coding, you’ve got to be curious and a comfortable with the process of gathering and understanding data. “A lot of the work is the journey,” said Neimeth. “The journey is the analysis, the exploration, the hypotheses, the testing.”

“If you want to get into data science, you need to think: Are you going to be happy on the journey as well as at the destination?”