Neil Perry is Content Director for Outlook Publishing.
- Content Director
Share
Blue Yonder has unveiled a new AI “Model Training Factory” developed with NVIDIA, aimed at creating specialised supply chain agents capable of automating warehouse, logistics and planning decisions at scale.
Blue Yonder targets autonomous supply chain execution with specialised AI agents
Blue Yonder has announced the launch of its Model Training Factory, a new AI development framework created with NVIDIA to accelerate deployment of specialised AI agents across supply chain operations.
Unveiled at the company’s ICON customer conference, the system is designed to fine-tune and test domain-specific supply chain models capable of supporting autonomous decision-making across warehouse management, transportation, inventory planning, merchandising and network operations.
The initiative combines NVIDIA Nemotron open-source models and NeMo AI tools with Blue Yonder’s operational supply chain data and workflow expertise.
Supply chain complexity driving next phase of enterprise AI
Blue Yonder said growing supply chain complexity, operational volatility and the need for low-latency decision-making are increasing demand for more specialised AI systems.
According to the company, enterprises are moving beyond general-purpose AI assistants toward teams of domain-specific agents capable of analysing conditions, using tools and executing workflows alongside human operators.
The economics of scaling large AI models across operational environments are becoming increasingly challenging, particularly as demand for AI inference capacity rises, and Blue Yonder said its model factory addresses this through a hybrid architecture that combines frontier AI models with smaller, workflow-specific supply chain models optimised for speed, precision and lower operating cost.
Purpose-built AI models trained for operational supply chain workflows
The Model Training Factory is designed as a repeatable system for building AI agents trained on supply chain-specific decisioning processes and models are fine-tuned to execute complex, multi-step workflows while being evaluated against strict performance criteria before deployment.
The company said the models are trained using synthetic data rather than customer data.
“Supply chain has always been an AI domain. Our research into how agentic models perform on real warehouse and planning decisioning is exactly why we know where frontier models hit a wall,” said Duncan Angove, CEO of Blue Yonder.
“Working with NVIDIA, we’re building owned intelligence, not rented intelligence—supply chain models trained on the workflows, telemetry, and decision logic that actually run a warehouse or a planning system. This isn’t a one-off fine-tuned model. It’s a factory, and it produces purpose-built agents at the speed, precision and cost the autonomous supply chain demands.”
Warehouse operations become first deployment focus
The first production models will focus on warehouse management workflows, including:
Allocation shortages
Inventory exceptions
Due-time urgency
Inventory management across yard and receiving trailers
These are high-frequency operational decisions where speed and accuracy directly affect on-time performance, inventory availability and order cycle times.
AI agents could continuously evaluate hundreds of operational trade-offs in seconds across warehouse environments where human operators typically assess only a limited number of variables.
NVIDIA partnership underpins model development infrastructure
The Model Training Factory is built using NVIDIA’s agentic AI technology stack, including Nemotron open-source models, the NVIDIA NeMo Agent Toolkit and NVIDIA AI Enterprise infrastructure.
The range of Nemotron model sizes allows AI systems to be matched to specific operational requirements, from compact warehouse optimisation models to larger planning and orchestration systems.
“The next phase of enterprise AI for supply chains requires specialized, affordable and accurate domain-trained agents that can operate within the workflows that run a business,” said Azita Martin, Vice President and General Manager, Retail and CPG, NVIDIA.
“Blue Yonder is leveraging NVIDIA Nemotron, the NVIDIA NeMo Agent Toolkit and NVIDIA AI Enterprise to build a Model Training Factory that fine-tunes models with proprietary supply chain data, enabling them to build agentic AI systems for some of the world’s largest and most complex supply chains.”
Operational expertise positioned as long-term competitive advantage
The factory model is designed to convert operational supply chain expertise into reusable AI training signals that can scale across workflows and industries.
The company identified its combination of workflows, telemetry, subject matter expertise, evaluations and retraining processes as a core differentiator for building specialised supply chain AI systems, and the first AI models developed through the factory are expected to enter customer production environments later this year through Blue Yonder Cognitive Solutions.
This article was produced by the editorial team at Supply Chain Outlook and published as part of the Outlook Publishing global network of B2B industry magazines.
Outlook Publishing delivers industry insights, company stories, and sector coverage across supply chains, manufacturing, mining, construction, healthcare, food production, and sustainability.
Supply Chain Outlook provides ongoing coverage of organisations and developments shaping the global logistics and supply chain sector.
