Data Minds: Jai Bansal — Data Scientist at Red BullData Minds is a series that profiles professionals working with data.
In this series, you’ll learn about their story, their day-to-day, as well as tips and advice for others.
Andrei LyskovBlockedUnblockFollowFollowingMar 18Red Bull has the highest market share of any energy drink company in the world, with 6.
79 billion cans sold in 2018 alone.
As you can imagine, this means that they generate a lot of data.
From images that drivers upload to verify delivery, to data on business operations.
Jai Bansal has been a data scientist at Red Bull for the past 2 and a half years, with 2 years of previous experience at a startup.
In this article, we talk about how he got into data science, what it’s like to be at a company with a budding data culture, the importance of spending 50% of your time learning, and general tips for others looking to break into data science.
From Economics to Data ScienceWhen Jai was doing his masters in Economics at USC in 2010, the term data science hadn’t taken off yet.
The term was only coined two years previously (2008), by DJ Patil and Jeff Hammerbacher, early data science leads at LinkedIn and Facebook.
While in school, Jai had a broad set of interests.
He knew that he enjoyed Math and had experience with R.
But outside of that, he didn’t have the practical data science skills that you might expect today.
That soon changed when he landed his first job out of school.
At the time, he knew he didn’t want to continue in academia, so he found and interviewed for a data related role at a startup.
Instead of waiting for the ‘perfect’ role, he realized it’s better to start working and getting industry experience, sooner rather than later.
Forced to learn all the industry tools to be effective, he soon found himself drinking from a fire hose.
While working at this startup wasn’t always the best experience, it did allow him to cut his teeth and get valuable experience.
He made sure to be proactive about taking on a wide array of projects at his company, to give him the breadth of exposure that he would need later in his career.
After the first two years there, he realized it was time for a change.
He interviewed and joined Red Bull as one of their first few data scientists in North America.
Data Science At Red BullSince the data science team at Red Bull was still in its early stages, it wasn’t as rough of a transition coming from a startup environment.
The nascent data culture gave the team a lot of opportunities to be nimble and experiential.
Some of the early projects he had a chance to work on ranged from tinkering with convolutional neural networks to building predictive models to help departments plan more effectively.
One of the earliest lessons that Jai received was related to solving the right problems.
He shared that spending too much time on projects without getting stakeholder feedback or buy-in, can often lead to a huge waste of resources.
He advocated for a lean-startup approach in iterating quickly and getting feedback before investing additional time into projects.
Another component of Jai’s work consists of working with key departments to find and select relevant projects.
Figuring out what motivates business stakeholders (hint: cutting costs, growing revenue and increasing profit), as well as working collaboratively with them to identify strategic objectives worth pursuing, is a big part of what he does.
This is especially true within a company with an early data culture, where stakeholders may not always immediately see the value of data scientists.
Jai also contributes to the data evangelism efforts within his company.
He runs their R analytics training and contributes to an internal newsletter.
Depending on the project, and what phase they’re in, he spends 70% of his time coding, and another 30% doing outreach or presenting his results.
This is then further broken down into 30% ad-hoc requests (something like a SQL query), 40% individual projects, and 30% group project with other data scientists on his team.
He codes in SQL, R, or Python depending on the project.
Advice for OthersJai’s biggest advice for people looking to break into data science is to love learning and to make sure you’re always learning on the job.
Data science is a field that’s changing, and so it’s important to stay on top of the latest developments.
His personal goal is to spend 50% of his work time on a project that pushes him to learn, and the other 50% on repetitive work he’s done before.
Working with your manager, and coming up with a long-term learning plan is another great piece of advice to make sure that you’re investing in your future development.
Besides that, Jai stressed the importance of non-technical skills.
Being able to communicate effectively with non-technical stakeholders goes a long way to building trust.
Nothing makes someone tune out quicker than feeling overwhelmed from technical jargon.
That said, it’s also important to take the time to educate your stakeholders.
The ability to explain the underlying concepts of what powers your predictive models can be a great way to lower the resistance someone may have to adopting it.
Finally, it’s important to remember that you and your data science team are part of a greater whole.
Putting the business first, and understanding the pain of your stakeholders will go a long way in creating an environment of trust and collaboration.
While it may be more intellectually stimulating to work on a deep learning model, it may, in fact, be more useful to start with a simpler model and build from there.