The cognitive and agentic supply chain isn't a roadmap anymore: customers are living it.
-1.png?width=800&height=600&name=image%20(7)-1.png)
For three consecutive years, Blue Yonder's ICON conference has told a progressive story. At ICON 2024 in Dallas, CEO Duncan Angove unveiled the cognitive vision. A new class of AI-native solutions built on a rebuilt cloud platform, backed by over $2 billion in R&D. At ICON 2025, the company launched its first generation of domain-specific AI agents and the cognitive planning suite, laying the architectural groundwork for agentic supply chain operations. Both years carried the energy of builders laying a foundation. The offerings were ambitious and technically impressive, but still largely about what was coming.
ICON 2026 in San Diego was different. The customers took center stage, and they were talking about more than pilots. They were sharing real operational results from cognitive and agentic deployments across retail, manufacturing, and logistics. The narrative shifted from "here's what we're building" to "here's what we're getting from it."
ICON 2026: Customers Take the Stage
Blue Yonder put its customers front and center, and the stories spanned industries and geographies.
Sainsbury's CEO Simon Roberts described how the retailer's end-to-end supply chain transformation has driven product availability to 98% across 30,000 SKUs, producing the retailer's highest volume market share in a decade. He highlighted Sainsbury’s ‘comeback’ and credited Blue Yonder with helping them get there.
Crate & Barrel CTO Jason Booth made the case for platforms over point solutions and shared results from the retailer's Bring It Home strategy. This included reduced peak-season backorders, improved online conversion, and increased fulfillment efficiency. He's now moving into agents for order health and inventory health to identify emerging risks before they reach customers.
Knauf, one of the world's largest building materials manufacturers with 300+ production sites across 90 countries, shared its ambition for a fully automated end-to-end supply chain by 2032. They are targeting 80% touchless orders and will be bringing in external partners to provide a singular view (the network) for enhanced efficiency. They are only three months live on cognitive planning and planners are already shifting from data reconciliation to judgment-intensive work.
Australia Post Group CEO Paul Graham described a 217-year-old organization running 12,000 vehicles across 12.8 million delivery points with no central brain. TMS with Blue Yonder will become that coordinating intelligence, with same-day delivery expectations revised from 5% to 20% of volume by 2030.
In an era of AI hype, it was refreshing to hear operational transformations at genuine enterprise scale.
The Agent Is the App
The philosophical anchor of ICON 2026 was Angove's declaration that Blue Yonder stopped selling anything other than its cognitive portfolio at the start of this year. Every deployment is now, from day one, an agentic deployment.
The phrase he returned to throughout: "the agent is the app." Agents build context over time, becoming operationally fluent with every interaction. The vision is that all applications will become headless and open, with agents generating custom UX by user, adapting to each person's role, context, and workflow. Personally, the hangover from the term ‘headless’ in the eCommerce world may create some friction, but the sentiment is solid.
The ‘agent is the app’ narrative extends across the portfolio. Agentic Merchandising leverages networked planograms for real-time co-creation between retailers, suppliers, and store teams, providing a single version of the truth on the Blue Yonder Network. Agentic Commerce addresses how AI-driven shifts in product discovery are fundamentally changing demands on order management.
Partners Contribute to the Narrative
Blue Yonder added to its partnership with NVIDIA and announced the Model Training Factory, a repeatable system for fine-tuning supply chain-specific AI models on NVIDIA's Nemotron open-source models and NeMo Agent Toolkit. The core of the solution is offering domain-specific models trained on supply chain workflows that outperform generic frontier models in use cases like short-term warehouse utilization. Models are trained on synthetic data, and the pipeline produces a continuous stream of increasingly capable, domain-trained agents.
Microsoft's Judson Althoff, CEO detailed how Blue Yonder uses Azure AI Foundry to design, test, and deploy personalized agents that automate planning, forecasting, and execution workflows, while its supply chain-optimized models run on Azure's high-performance AI infrastructure.
The Syndigo partnership addresses a different but equally critical layer: product data quality. Another key theme of this year’s ICON was the Data Quality foundation that is necessary for any agentic deployment. The partnership integrates Syndigo's enriched, GS1-aligned product content, validated attributes, standardized images, and accurate dimensions directly into Blue Yonder's supply chain and space planning workflows. The claim is that this partnership will generate things like higher-confidence planograms built on trusted data, resulting in faster resets, fewer exceptions, and measurable improvements in store execution compliance.
Frictionless Adoption: Turning the Customer Lifecycle into a Product Feature
Blue Yonder announced the Frictionless Outcomes Manifesto. Although a bit ‘Jerry Maguire-ish’ in name, it is a commitment to automating the software implementation lifecycle itself. The traditional deployment model of statements of work and time-and-materials contracts is what Angove calls "linguistic fossils" from the industrial era.
The numbers the team were throwing out, however, were compelling: WMS migration initiation reduced by 87%, from eleven weeks to seven days. A major planning use case migrated in 72 hours, configuration becomes inference. Data ingestion dropped from two weeks to eight hours.
This requires a change in project management processes across the industry. As innovation costs come down, innovation velocity goes up. This could have serious implications for the entire SI ecosystem, which Angove suggested is about to become a product feature.
Embedding for Success: The FDE
All agentic deployments use Forward Deployed Engineers (FDEs). These are Blue Yonder's own people embedded directly inside customer operations alongside planners, supervisors, and logistics teams. They translate tribal knowledge and operational behavior into agentic intelligence.
By placing its own engineers inside the outcome rather than handing it to a systems integrator, Blue Yonder is betting the journey to cognitive and agentic goes more smoothly when the vendor has skin in the game. This model also generates the operational telemetry that makes agents smarter for every subsequent deployment. The goal, as the Blue Yonder team framed it, is to monetize the network. The more customers deploy, the more the collective intelligence improves.
The Incisiv Take
ICON 2026 marked a turning point. The two prior years of foundation-building have crossed from roadmap to reality. Customers across retail, manufacturing, logistics, and defense are running cognitive and agentic solutions in production and getting measurable value.
What distinguishes Blue Yonder's approach is not just the technology, but the go-to-market model wrapped around it. The Frictionless Outcomes Manifesto, the forward deployed engineers, the commitment to making deployment itself a product feature signal a company that understands the hardest part isn't building the AI. It's changing how organizations operate.
The question for the industry is no longer whether cognitive and agentic supply chain solutions can deliver value. The question is how quickly organizations are willing to change their operating models to capture it.
