'Data Science And AI Aren’t Silver Bullets’: Kroger’s First Chief Data And AI Officer On Closing The Impact Gap

Chief Data and AI Officers are emerging as central figures in the race to turn potentially decades of rich operational and customer information into business impact.  

Earlier this month, Kroger named Milen Mahadevan, currently president and CEO of its 84.51° subsidiary, as also the company’s first CDAO. The move comes as Kroger shifts its strategy from standalone AI tools, to building foundational capabilities that orchestrate work, improve decision making, and create more connected experiences. Over the past year, Kroger expanded use of its AI-powered digital shopping assistant, rolled out Sage, an internal AI virtual assistant used by over 150,000 associates, and launched collaborations with leading AI labs.

As CDAO, Milen oversees all of Kroger’s data and AI teams, and works closely with business leaders from merchandising, supply chain, marketing, store, and eCommerce operations to instill more focus and accountability across the company’s AI investments.

“AI is no longer experimental, but central to how we operate, scale impact, and create value,” Milen told CIO News. It’s “an opportunity to connect Kroger’s unique data assets, our science and technology capabilities, and our operating model in a way that can reshape how we work and how we serve customers.”

A chemical engineer by training, Milen parlayed his strong background in math, physics, and problem-solving into a career in the technology industry. His journey began at dunnhumby, a customer data science and analytics provider that, at the time, was partnering closely with Kroger on a joint venture. Over a nearly 15-year career, Milen had a role in data science and engineering, application development, and client engagement. The experience ended up changing his philosophy around data and AI, and would ultimately lead him to his current role.

“It’s easy to create impressive models or experiments, but what truly matters is impact — how these capabilities are used to drive meaningful business outcomes,” Milen said.

In 2015, dunnhumby USA became 84.51°, a wholly owned subsidiary of The Kroger Co., focused on embedding data science deeply within the company's business lines. Milen rose through the ranks, eventually becoming president and CEO. Now, with CDAO added to his title, Milen is helping Kroger fortify itself amid the ongoing AI revolution.

“Data science and AI aren’t silver bullets; they have to get under the skin of the real problem trying to be solved,” Milen said. “I've seen the power of Kroger's data from every angle: how it fuels insights for our collaborators, personalizes the customer experience, and is already driving results that allow us to be more competitive, improve shrink, deliver faster fulfillment and amplify the talents of our associates.”

We connected with Milen to discuss tapping into Kroger’s data treasure trove, moving beyond AI pilots, and why people are the centerpiece to any successful AI strategy.

The data advantage

With a long history in data science and analytics, Kroger isn’t starting the AI race from behind.

As third-party signals became less reliable over the last few years, 84.51° turned its attention to transforming Kroger’s proprietary, first-party data into actionable intelligence and growth — work that now serves as the foundation for its AI efforts. And with online sales a $16 billion business and growing quarter-over-quarter, it’s more data, more impressions, and ultimately more value for Milen and his team to deliver to Kroger and its partners.

“In a world where everyone has access to the same foundational models, proprietary data is the ultimate differentiator.” he said. “Kroger’s data advantage isn’t just its size, but the depth and quality of signal: real purchase behavior across channels, price points, and occasions — strengthened by our loyalty ecosystem.”

But as AI advances at a pace few expected, the challenge for enterprises like Kroger is deploying the technology with both agility and purpose. For Milen, it all comes down to one question: How do you become an enterprise that can operate through AI?

“Answering that requires more than pilots—it requires leadership, focus, strong governance, sustained platform investment and a clear operating model to scale impact across the business,” he said.

This unique blend of skillsets is what led Kroger to create the CDAO role. Data and AI teams now operate under a single lead and vision: develop next-generation agentic capabilities that maximize the company’s data assets to reinvent how Kroger operates as an enterprise and engages with customers.

“We believe AI is becoming a core operating layer of the business — not something you simply bolt onto the side,” Milen said. “The organizational and foundational investments we’re making now are about ensuring we can capture that opportunity in a disciplined, responsible and differentiated way.”

People at the heart

Hidden beneath an AI agent’s simple, natural language interfaces is a complex ecosystem of different components. Enterprise-ready systems require well-governed data that flows reliably to every endpoint. The technology also has to be integrated into the everyday experience of associates and employees. And it all must be done in a way that respects and protects customer privacy.

“The AI model is often the easiest part. The hard work is data preparation, system integration, change management, and continuous iteration based on real feedback,” Milen said.

And while buying AI tools is easy, actually getting the workforce to embrace them in meaningful ways is much harder: “Technology doesn’t transform a company — people do,” Milen said.

That mindset lives at the heart of Sage, a virtual AI assistant for associates. Designed specifically to solve real-world pain points, employees are rapidly infusing the tool into their daily work. According to Milen, Sage is already helping to eliminate administrative and manual burdens on associates, as well as deliver operational improvements, like reducing the amount of lost, stolen, or damaged product. And importantly, Sage is giving employees the opportunity to focus on the customer experience.

“Associates have embraced it faster than we expected, which reinforces that when you design AI with end users, adoption follows naturally,” said Milen. “We’re still early in what we can achieve, and building capability across the organization is a priority — but this won’t happen because of a single role or team. It happens when leaders and associates across the company are aligned around using AI to reinvent how we operate.”