How Walmart built one of the world’s largest enterprise AI operations

Walmart has quietly assembled one of the world’s most sophisticated enterprise AI operations, managing thousands of use cases across its 4,700 stores and 255 million weekly customers. Rather than deploying AI as an experimental add-on, the retail giant has fundamentally restructured how it approaches artificial intelligence—treating trust not as a compliance afterthought, but as a core engineering requirement.

During VB Transform 2025, Desirée Gosby, Walmart’s VP of Emerging Technology, revealed how the company operationalizes AI at unprecedented scale. Her insights offer a rare glimpse into how enterprises can move beyond pilot programs to achieve meaningful AI transformation. “We see this as a pretty big inflection point, very similar to the internet,” Gosby explained during the session. “It’s as profound in terms of how we’re actually going to operate, how we actually do work.”

The stakes couldn’t be higher. With millions of daily transactions flowing through a complex ecosystem of stores, suppliers, and digital platforms, Walmart’s AI strategy provides a tested framework for any enterprise grappling with large-scale artificial intelligence deployment.

Four distinct stakeholders drive AI architecture

Walmart’s approach rejects the typical one-size-fits-all AI platform in favor of targeted solutions designed for specific user groups. This stakeholder-focused architecture acknowledges that store associates managing inventory face entirely different challenges than merchants analyzing regional buying patterns.

The company segments its AI tools across four primary groups. Customers interact with Sparky, Walmart’s natural language shopping assistant that helps navigate product selection. Field associates receive inventory optimization tools and workflow management systems designed for store operations. Merchants access sophisticated decision-support systems for category management and trend analysis. Sellers get business integration capabilities that streamline their operations within Walmart’s ecosystem.

“And then, of course, we’ve got developers, and really, you know, giving them the superpowers and charging them up with, you know, the new agent tools,” Gosby explained. This developer-focused approach ensures that technical teams can build and customize AI solutions rapidly across the organization.

“We have hundreds, if not thousands, of different use cases across the company that we’re bringing to life,” Gosby revealed. This scale demands architectural discipline that most enterprises lack, requiring systematic approaches to prevent the technical debt that typically accompanies rapid AI expansion.

The segmentation strategy drives adoption through relevance rather than mandate. When tools directly address specific operational pain points, users embrace them naturally rather than requiring extensive change management programs.

Trust emerges through value delivery, not training programs

Walmart discovered that sustainable AI adoption happens when systems deliver immediate, tangible benefits rather than through mandatory training sessions that employees often question. Gosby illustrated this principle through her mother’s shopping evolution—from weekly store visits to COVID-era deliveries to predictive commerce.

“She’s been interacting with AI through that whole time,” Gosby explained. “The fact that she was able to go to the store and get what she wanted, it was on the shelf. AI was used to do that.” Each interaction provided clear value without requiring her mother to understand the underlying technology.

This progression demonstrates Walmart’s broader predictive commerce vision. “Instead of having to go weekly, figure out what groceries you need to have delivered, what if it just showed up for you automatically?” Gosby asked. The concept represents AI’s ultimate goal: anticipating customer needs before they’re explicitly expressed.

“If you’re adding value to their lives, helping them remove friction, helping them save money and live better, which is part of our mission, then the trust comes,” Gosby stated. This principle applies equally to internal users. When AI genuinely improves associates’ work efficiency and effectiveness, adoption happens organically.

The trust-through-value approach eliminates the resistance typically associated with AI implementations. Rather than fighting skepticism, Walmart builds confidence through proven results.

Fashion cycles compress from months to weeks through AI acceleration

Walmart’s Trend to Product system exemplifies how AI transforms traditional retail operations. The platform synthesizes social media signals, customer behavior data, and regional purchasing patterns to dramatically accelerate product development timelines.

“Trend to Product has gotten us down from months to weeks to getting the right products to our customers,” Gosby revealed. This compression fundamentally alters retail economics by enabling response to real-time demand rather than historical forecasting.

The operational impact extends far beyond speed. Faster product cycles mean improved inventory turnover, reduced markdown exposure, and enhanced capital efficiency. Walmart maintains its price leadership position while matching any competitor’s speed-to-market capabilities across high-velocity categories.

This acceleration demonstrates AI’s ability to transform entire business models. Traditional retail planning, which relied on seasonal buying cycles and historical data, gives way to dynamic systems that respond to emerging trends within weeks rather than planning seasons.

Model Context Protocol enables scalable agent architecture

Walmart’s technical approach to AI orchestration draws from hard-won lessons in distributed systems architecture. The company uses Model Context Protocol (MCP)—a standardized framework for how AI agents interact with existing business systems—to avoid the integration chaos that typically derails enterprise AI initiatives.

“We break down our domains and really looking at how do we wrap those things as MCP protocol, and then exposing those things that we can then start to orchestrate different agents,” Gosby explained. This strategy transforms existing infrastructure rather than requiring complete system replacement.

The architectural philosophy reflects deeper organizational learning. “The change that we’re seeing today is very similar to what we’ve seen when we went from monoliths to distributed systems. We don’t want to repeat those mistakes,” Gosby stated. The reference points to the painful transition many enterprises experienced moving from single, large applications to distributed service architectures.

Gosby outlined the practical execution requirements: “How do you decompose your domains? What MCP servers should you have? What sort of agent orchestration should you have?” At Walmart, these represent daily operational decisions rather than theoretical planning exercises.

“We’re looking to take our existing infrastructure, break it down, and then recompose it into the agents that we want to be able to build,” Gosby explained. This standardization-first approach enables flexibility while maintaining system stability. Services built years ago now power modern AI experiences through proper abstraction layers.

Decades of merchant expertise becomes enterprise intelligence

Walmart’s competitive advantage extends beyond technology to the systematic capture of human knowledge accumulated over decades. The company transforms the expertise of thousands of category specialists into accessible AI capabilities that digital-first retailers cannot replicate.

“We have thousands of merchants who are excellent at what they do. They are experts in the categories that they support,” Gosby explained. “We have a cheese merchant who knows exactly what wine goes or what cheese pairing, but that data isn’t necessarily captured in a structured way.”

AI operationalizes this institutional knowledge. “With the tools that we have, we can capture that expertise that they have and really bring that to bear for our customers,” Gosby said. The practical application becomes clear in customer interactions: “When they’re trying to figure out, hey, I need to throw the party, what kind of appetizers should I have?”

This strategic advantage compounds over time. Decades of merchant expertise become accessible through natural language queries, creating a knowledge base that algorithms cannot synthesize independently. Walmart’s 2.2 million associates represent proprietary intelligence that provides sustainable competitive differentiation.

The approach transforms individual expertise into enterprise assets. Knowledge that previously resided in specific employees’ minds now enhances customer experiences across all touchpoints.

New metrics measure autonomous system success

Traditional business metrics fail when AI agents handle complete workflows rather than supporting human-driven processes. Walmart pioneers measurement systems designed for autonomous operations that extend beyond conventional conversion tracking.

“In an agentic world, we’re starting to work through this, and it’s going to change,” Gosby said. “The metrics around conversion and things like that, those are not going to change, but we’re going to be looking at goal completion.”

The shift reflects operational reality in AI-driven environments. “Did we actually achieve what is the ultimate goal that our associate, that our customers, are actually solving?” Gosby asked. This question reframes success measurement from process compliance to problem resolution.

“At the end of the day, it’s a measure of, are we delivering the benefit? Are we delivering the value that we expect, and then working back from there to basically figure out the right metrics?” Gosby explained.

This metrics evolution acknowledges that AI systems often accomplish objectives through paths that differ from traditional customer journeys. Success measurement must adapt to autonomous decision-making rather than human-guided processes.

Enterprise framework scales beyond retail

Walmart’s systematic approach provides actionable intelligence for enterprise AI deployment across industries. The company’s operational framework, validated through millions of daily transactions, offers principles that extend far beyond retail operations.

Build architectural discipline from day one. Walmart’s experience transitioning from monolithic to distributed systems provided crucial lessons for AI deployment. The key insight: establish proper foundations before scaling to prevent expensive rework later. Define systematic approaches that accommodate growth without architectural debt.

Match solutions to specific user needs. Generic AI platforms consistently fail because they ignore operational reality. Store associates require different capabilities than merchants, who need different tools than suppliers. Walmart’s targeted approach drives adoption through relevance rather than mandate.

Establish trust through proven value. Begin with clear wins that deliver measurable results before expanding scope. Walmart progressed from basic inventory management to predictive commerce step by step, with each success building credibility for the next phase.

Transform employee knowledge into enterprise assets. Decades of specialist expertise exist within most organizations. Walmart systematically captures merchant intelligence and operationalizes it across 255 million weekly customer interactions. This institutional knowledge creates competitive advantages that algorithms cannot replicate from scratch.

Measure what matters in autonomous systems. Traditional conversion metrics miss the point when AI handles entire workflows. Focus on problem resolution and value delivery rather than process tracking. Walmart’s metrics evolved to match operational reality rather than forcing AI into traditional measurement frameworks.

Standardize before complexity overwhelms systems. Integration failures destroy more AI projects than poor algorithms. Walmart’s protocol decisions prevent the chaos that typically derails enterprise initiatives. Structure enables speed by providing consistent frameworks for rapid development.

“It always comes back to basics,” Gosby advised. “Take a step back and first understand what problems do you really need to solve for your customers, for our associates. Where is there friction? Where is there manual work that you can now start to think differently about?”

Implications extend across industries

Walmart’s framework applies to any enterprise managing complex stakeholder relationships. Financial services organizations balancing customer needs with regulatory requirements face similar multi-stakeholder challenges. Healthcare systems coordinating patient care across providers encounter comparable complexity. Manufacturers managing intricate supply chains deal with analogous coordination problems.

“It’s permeating everything it is that we do,” Gosby explained. “But at the end of the day, the way that we look at it is we always start with our customers and our members and really understanding how it’s going to impact them.”

The methodology prioritizes problem resolution over technology deployment. “Our customers are trying to solve a problem for themselves. Same thing for our associates,” Gosby stated. “Did we actually solve that problem with these new tools?”

This focus on outcomes rather than implementation drives measurable results. Walmart’s scale validates the approach for any enterprise ready to move beyond pilot programs toward systematic AI transformation. The company demonstrates that enterprise AI succeeds through engineering discipline, stakeholder-focused design, and relentless attention to value delivery rather than technological sophistication alone.