There’s no doubt that data scientists are in high demand right now. Companies are looking for people who can help them make sense of all the data they’re collecting and use it to make better decisions.

Being a data scientist is a great way to start or further your career. It’s a field rising, and there are many opportunities for those with the right skills. We talked with Ken Jee, Head of Data Science at Scouts Consulting Group, about how to build a career in data science.

With a goal-oriented approach to problem solving, data science evangelist Ken Jee is admired for his work in the field. He is the Head of Data Science at Scouts Consulting Group, and creates and shares content via his podcast, website, YouTube channel, and 365 DataScience course offering; frequently contributes to Kaggle; and is a Z by HP global ambassador. He recently helped design data science challenges featured in “Unlocked,” an interactive film from Z by HP. The film and companion website present data scientists with the opportunity to participate in a series of problem-solving challenges while showcasing the value of data science to non-technical stakeholders, with a compelling narrative. We spoke with Jee about how he built a successful career as a data scientist.

What is your background and how did you get started in data science?

All my life, I played competitive sports, and I played golf in college. One of the ways that I found I could create a competitive edge was by analyzing my data and looking at the efficiencies that could be created by better understanding my game, and allocating time towards practice more effectively. Over time I became interested in the data of professional athletes, particularly golfers, so I started to analyze their performance to predict the outcome of events. I tried to play golf professionally for a bit, but it turns out I am better at analyzing data than playing the game itself.

What advice do you give young people starting out or wanting to get into the field?

If they’re just starting out learning data science, I recommend that they just choose a path and stick to it. A lot of times people get really wrapped up in whether they’re taking the right course and end up spinning their wheels. Their time would be better spent just learning, whatever path they take. I will also say that the best way to land a job and get opportunities is by creating a portfolio by doing data science. Create or find data, whether it’s on Kaggle or from somewhere else, like the “Unlocked” challenge, show your work to the world, get feedback and use that to improve your skills.

“Unlocked” is a short film that presents viewers with a series of data science challenges, that I along with other Z by HP Data Science Global ambassadors helped to design. There are challenges that involve data visualization using environmental data; natural language processing or text analysis using a lot of synthesized blog posts and internet data; signal processing of audio information; and computer vision to analyze pictures, along with accompanying tutorials and sample data sets. We wanted to highlight a variety of things that we thought were very exciting within the domain.

There’s a lot of fun in each of these challenges. We’re just really excited to be able to showcase it in such a high production value way. I also think that the film itself shows the essence of data science. A lot of people’s eyes glaze over when they hear about big data, algorithms and coding. I jump out of bed in the morning happy to do this work because we see the tangible impact of the change that we’re creating, and in “Unlocked,” you’re able to follow along in an exciting story. You also get to directly see how the data that you’re analyzing is integrated into the solutions that the characters are creating.

How has technology opened doors for you in your career?

I would argue that technology built my entire career, particularly machine learning and AI tech. This space has given me plenty to talk about in the content that I create, but it has also helped to perpetuate my content and my brand. If you think about it, the big social media companies including YouTube all leverage the most powerful machine learning models to put the right content in front of the right people. If I produce content, these algorithms find a home for it. This technology has helped me to build a community and grow by just producing content that I’m passionate about. It is a bit meta that machine learning models perpetuate my machine learning and data science content. This brand growth through technology has also opened the door for opportunities like a partnership with Z by HP as a global data science ambassador. This role gives me access to and the ability to provide feedback on the development of their line of workstations specifically tuned to data science applications–complete with a fully loaded software stack of the tools that my colleagues and I rely on to do the work we do. Working with their hardware, I’ve been able to save time and expand my capabilities to produce even more!

What educational background is best suited for a career in data science?

I think you have to be able to code, and have an understanding of math and programming, but you don’t need a formal background in those areas. The idea that someone needs a master’s degree in computer science, data science or math is completely overblown. You need to learn those skills in some way, but rather than looking at degrees or certificates, I evaluate candidates on their ability to problem solve and think.

One of the beautiful things about data scientists is that they come from almost every discipline. I’ve met data scientists from backgrounds in psychology, chemistry, finance, etc. The core of data science is problem solving, and I think that’s also the goal in every single educational discipline. The difference is that data scientists use significantly more math and programming tools, and then there’s a bit of business knowledge or subject area expertise sprinkled in. I think a unique combination of skills is what makes data science such an integral aspect of businesses these days. At this point, every business is a technology company in some respect, and every company should be collecting large volumes of data, whether they plan to use it or not. There’s so much insight to be found in data, and with it, avenues for monetization. The point is to find new opportunities.

What’s an easy way to describe how data science delivers value to businesses?

At a high level, the most relevant metric for data science in the short term is cost savings. If you’re better able to estimate how many resources you’ll use, you can buy a more accurate number of those resources and eventually save money. For example, if you own a restaurant and need a set amount of perishable goods per day, you don’t want to have excess inventory at the end of the week. Data science can be used to very accurately predict the right quantity to buy to satisfy the need and minimize the waste, and this can be on-going and adjusted for new parameters. Appropriate resourcing is immensely important, because if you have too much, you’ll have spoilage, and too little, you’ll have unhappy customers. It’s a simple example but when your sales are more accurate, even by a small percentage, those savings compound. At the same time, the data science models get better, the logic improves, and all these analytics can be used for the benefit of the business and its profitability.  

Is being a data scientist applicable across industries?

You can have success as a data scientist generalist, where you bounce across different subject area expertise and industries, like finance, biomedical, etc.; you just have to be able to pick up those domains relatively quickly. I also think that if you’re looking to break into data science from another field, the easiest path for you would be to do data science in that field. It all sort of depends on the nature of the problems you would like to solve. There are verticals where subject area expertise is more important, maybe even more so than data skills, like for sports and you need to understand a specific problem. But generally, someone could switch between roles.

Any final notes?

I’m a huge believer of setting goals and accountability. A good goal is measurable, you control the outcome, and set a time constraint. Once you’ve set your goal, write it down or tell people about it. Also, never forget that learning is a forever journey.

Previous post

Is DataOps more than DevOps for data?

Next post

Machine learning engineering: The science of building reliable AI systems