With data science growing in sophistication and applications every day, it may be surprising to learn that some of the market’s biggest boosts are coming from companies with a primary focus on standard retail, according to Esteban Arcaute (pictured), senior director and head of data science at @WalmartLabs.

The impact of data science on sales, and the incentive for those retailers to enhance their practices with data science, are forming an important partnership of commerce and science, Arcaute said. He spoke with Lisa Martin (@Luccazara), co-host of theCUBE, SiliconANGLE’s mobile live streaming studio, at the Stanford Global Women in Data Science Conference. (*Disclosure below.)

Arcaute discussed the role of data science in retail, noting the industry generates an enormous amount of data across a wide variety of interactions. For a company the size of Walmart, “A lot of ready-made solutions that are available for other companies, they’re just not viable for us,” he said.

The effort to predict customer satisfaction and other consumer behaviors comprises a significant chunk of retail data calculations, naturally, but a part of that — which is gaining attention as data science evolves — is centered on predicting what those customers will want in the future. Arcaute likened it to how in the ’90s, having a search engine actually work “was like a miracle.” But now, with search boxes commonly appearing as minor embedded elements of larger applications, the expectations placed on them have grown past that low, initial hurdle.

In the same way, expectations for what Walmart might be capable of achieving with the data it gathers are rising, and not just from similar companies with their own digital transformation, but from the customers who are providing that data. In its efforts to anticipate future services, @WalmartLabs is working to unite merchandising with data science, Arcaute explained, while also “[ensuring] that there’s never one side that’s working on something that the other cannot leverage. … We create our systems to deeply integrate the merchandising functionality.”

Human intelligence, machine learning

As the company tries to identify trends with both sides of that equation, machine learning is playing its own part in the application of @WalmartLabs’ data sciences. But as Arcaute noted, humans continue to play an important role even in machine learning, as they’ve found that “when we have a human in the loop, we’ve seen that the acceleration [of machine learning] is actually measurable.”

Arcaute also felt that drawing from the best talent available for those human insights was the best way to go, and a way that could only followed by giving equal consideration to both men and women. “I think that starting with women in data science, it’s a natural place to start. … There should be a way to funnel all of this talent into data science,” he said.

In Arcaute’s analysis, the continued vagueness of data science’s definition was an asset to it, in some ways, as it allowed for approaches that weren’t bound by a strict set of expectations and beliefs. As he shared, there were really only three big qualities that @WalmartLabs looked for in its potential data scientists, those being skills in analysis, software engineering and communication.

Along with those desired qualities, the ability to recognize and develop ideas was a highlighted asset, which led to discussion of @WalmartLabs’ use of hackathons to identify potential talent. “Great ideas come from anywhere, and hackathons are places where you see all these ideas bubbling,” Arcaute said. He felt that a similar atmosphere was in evidence at the WiDS conference, with a line of speakers he found “amazing,” drawn from “all types of industry … and academic journeys.”

Watch the complete video interview below, and be sure to check out more of SiliconANGLE and theCUBE’s coverage of the Stanford Global Women in Data Science (WiDS) Conference. (*Disclosure: TheCUBE is a media partner at the conference. Neither Stanford nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo by SiliconANGLE