Walmart employs agent AI with major machine learning platform upgrades and developer “Super Agent”

Walmart Inc. offers major agent upgrades to the Element Machine Learning Platform, adding “super agents” for software development to its growing portfolio of semi-autonomous software robots.

The retail giant is using today's Converge 2025 technology event to reinforce the message that everything is an agent. This is an artificial intelligence worker honeybee who takes action and operates with minimal supervision. Element, a cross-cloud machine learning platform, is refactored to coordinate the military of agents. Wibey is a new superagent designed to provide developers with the best tools, solve pipeline problems and generally increase productivity.

Rethinking retail

In an interview with Siliconangle of Sravana Karnati, executive vice president of Walmart's global technology platform, the agent said it was more than an upgrade. They are key to rethinking how Walmart builds and expands intelligent systems.

“We are in a pivotal moment in the evolution of AI,” he said. “The focus is shifting to building an intelligent system driven by agents that can operate independently inference, memory and context.”

Element extensions support autonomous systems with the ability to interact with memory, inference, and multiple tools and application programming interfaces. The platform is given a stateful architecture that allows agents to track what they do and, for example, what they intend. This means that agents can maintain context – not just a single interaction, but an entire long-term workflow that spans into systems and services.

Karnati said it is important for tasks such as identifying and fixing code compliance gaps, verifying modifications to the test environment, and issuing required pull requests.

Coordinating Agents

The plug-in and tool-call ecosystem allows agents to interact natively with external systems. Element supports standardized communication protocols, including new agent-to-agent agent and model context protocols, for coordination between agents and applications. API orchestration extends coordination across distributed environments.

Walmart said that these features combine to act as a platform for managing the full lifecycle of intelligent agents, allowing them to work from development and testing to deployment and monitoring. There are also observability tools that allow teams to track decision paths, inference steps, and performance metrics.

Karnati said the change would allow the elements to be able to handle the same way agents historically handled machine learning models. Developers can register agents in the agent catalog, evaluate their behavior with built-in guardrails, and reuse them across multiple projects.

Wibey is a centralized interface for interacting with Walmart's internal systems. This is the fifth in the growth of our portfolio of superagents targeting customers, associates and partners. Carnati said superagents are essentially coordinating task-specific agents' activities, with Walmart currently having around 200 people.

Built on top of the elements, Wibey interprets user intent and root commands into the appropriate tool, API, or agent. It provides a unified entry point for software engineers, architects and product managers to perform common tasks.

Tamming Tool

Walmart has been building large development tools for many years and sees it as a way to organize and apply Wibey. “Wibey is not a dashboard or a portal,” Karnati said. “This is a calling layer that understands developer goals and coordinates execution across Walmart's technology ecosystem.”

For example, Walmart has developed an agent that can scan the codebase for accessibility compliance issues, identify gaps, apply recommended fixes, and validate changes in a test environment. Another set of agents can monitor third-party libraries for outdated or vulnerable versions, upgrade, test compatibility, and automatically create pull requests.

Wibey integrates with popular development tools such as the command line interface, Slack, and Visual Studio. It also supports distributed development by allowing Walmart teams to register and publish their agents through shared protocols and contracts.

Karnati said one of the strengths used by Wibey's cores is to generate tailored starter kits for software projects. Developers no longer need to search the internal portal for templates and services. Instead, you can use Wibey to generate customized project frameworks based on prompts and existing contexts, reducing the time spent searching.

“Previously, developers had to manually search for the right API or template,” Karnati said. “Wibey can develop a starter kit that is very specific to your project. There's no need to make many changes. It's ready-made based on the questions you're asking, and we know the context in which it works.”

Karnati said the shift to the agent system lays the foundation for a new phase of AI adoption at Walmart. Walmart supports more intelligent and adaptive systems in every part of the business, as well as greater automation.

In software development, agents can handle tasks such as debugging deployment pipelines and detecting and correcting infrastructure anomalies. Customer-facing agents are already used to personalize conversations.

The impact of agents such as Wibey can surpass the impact of traditional generation AI tools, mainly thanks to inference capabilities and integration with core enterprise systems. Karnati said.

“This isn't just about code generation,” he said. “It's about making decisions, implementing actions and building a system that can improve over time.”

Photo: Wikimedia Commons