With a variety of domains within the field of data science, it can get quite confusing for new professionals to navigate. Sharath Gokula, Director of Data Sciences and Analytics, Walmart eCommerce, speaks with Tristan Spaulding, Head of Product, Acceldata, about the growth pathway for data science professionals and the organization’s approach to data science. 

Because the data science space is quite wide with overlapping roles, new data professionals need to network, read a lot, and understand their areas of interest, Gokula says. They can even diversify their skills once they are on the job. He gives an example of analytics professionals picking up analytics skills and then moving over to the data scientists’ area. In his case, Gokula has moved around, from core data science through analytics to leadership roles within analytics.

“Data science is really large; it goes right from analytics to optimization and causal research. It borders machine learning, therefore, there’s an engineering aspect to it. You need to have a good understanding of the business you are associated with. It’s really important for people trying to get into data science to have a good sense of what excites them within this. Understand what exactly your core areas of interest and skills are and then make a transition,” Gokula says.

One’s interest could be in the analytics area where they are making sure the right decisions are made or in the data engineering side where they are enabling data parsing and getting the data into the data space, he notes. It could also be in the machine learning area where they are building systems that are integrated into engineering, or areas such as search recommendation systems.

He adds, however, that certain skills and attributes are always useful, such as programming skills, understanding statistics, and being excited about the business and its mission. And, while awareness of all aspects of a variety of areas is good,  one does not need to be an expert in all the areas, Gokula emphasizes.

“Some roles within data science are very engineering heavy. Data engineering is more an engineering function than an analytics function. ML engineer sits somewhere between a statistician and a very strong engineer. And an analytics engineer is normally human facing, and they can do a lot of these different things but they don’t really specialize in an area,” he adds.

Breaking it down further, he says that it also depends on business needs and how the company transitions. For example, 80% of the journey can be completed with general skills post-transition, but there would be a need for further optimization that can only be performed by specialists.

In his current role at Walmart, Gokula is more engaged in the analytics area, so he is involved in experimentation and making sure there are no surprises with outlier detection. However, he stresses that it is the will of the organization that matters more than the individual teams.

“The leadership believes data is a key differentiator. We see focus right from leaders, asking for specific data before they make decisions,” Gokula continues. So, a lot of the optimization decisions are based on data and that really stems from the will of the organization and the leadership. The retail business and e-commerce touch customers at a very deep level. And we’re constantly looking at what is the best experience and what they mean for customers, and how the space works. All of this is not possible without data.”

Further discussing his role, Gokula says that the most fascinating areas are the challenges and the opportunities. A large organization like Walmart is constantly competing with several different groups along with the newer digital native or cloud-native organizations. These are the ones that lack the rich history and the background of Walmart.

Asked about the approach to tackling that sort of competition, Gokula says that it lies with the leadership and what they’re thinking about regarding the next development in the technology vectors in the market. “They are that forward-looking. We keep a close eye on what’s happening generally in the market,” he adds.

Gokula notes that customers’ needs change, which is probably the most important factor. And these customers need support. As an example, he says that e-commerce and technology and related factors were not in play 20-25 years ago, but they are an absolute necessity for customers today.

“People shopping on apps, for instance, are generally inquisitive, which represents a customer need. It has to be serviced by technology and the underlying data associated with it. So we keep an eye on all of this — on the competition, our customers, and where the general data market is heading,” he concludes.