Walmart introduces AI platform, 'super agents'
Walmart continues to develop its machine learning and agentic artificial intelligence capabilities.
The discount giant is in the middle of an ongoing technology transformation effort to redefine how retail operates at scale using AI, automation, and data intelligence. This includes the launch of an ML platform, called "Element," that is purpose-built to support agentic AI, which expands upon the prescriptive capabilities of generative AI to streamline enterprise workflows even further.
[READ MORE: How agentic AI is finding its place in retail]
Based on the open-source application management system Kubernetes, Element is designed to let users discover and reuse existing AI models; prepare data consistently; and build, test, and deploy new models rapidly while meeting compliance and governance standards.
Now, Walmart has introduced a suite of new capabilities designed specifically for agentic AI functionality, including:
- Stateful architecture that tracks what agents say, do and aim to achieve — managing both short-term and long-term memory.
- Agent-aware pipelines that propagate context and retrieve relevant information at the right time.
- Tool calling and plugin ecosystems that allow agents to interact with external systems and services natively.
- Standardized communication protocols to help ensure seamless coordination between agents and other entities.
- API orchestration that helps enable agents to perform complex tasks across distributed systems.
Element also provides observability into agent behavior, including decision paths, reasoning steps, and tool usage to help users monitor and measure custom metrics and dynamic behaviors to support continuous improvement of intelligent systems.
WIBEY ‘super agent’
Built on Element, Walmart’s WIBEY agentic AI "super agent" is an invocation layer that interprets agent developer intent and orchestrates execution across Walmart’s agentic ecosystem.
With WIBEY, a single prompt can scaffold a service, resolve a compliance issue, or fix a pipeline, without needing to know which system handles what. It transforms fragmented workflows into coordinated systems by embedding directly into developer tools, routing intelligently across domains, and supporting decentralized agent contribution via shared contracts.
AI agent builders, as well as non-builder personnel who use agents, can leverage WIBEY. To support intelligent orchestration at scale, WIBEY is built on a federated model where domain teams own their agents, while WIBEY helps make them discoverable and interoperable using protocols like MCP for discovery and access and A2A for delegation and chaining..
"By equipping developers with powerful tools like Element and WIBEY, we’re unlocking faster execution, greater autonomy, and a new standard for intelligent system design across Walmart," Sravana Karnati, executive VP, global technology platforms, Walmart Global Tech, said in a corporate blog post. "As we embrace agentic AI, we’re building a smarter Walmart — one that sets new standards for speed, precision, and service, and where intelligent systems work alongside us."
