Blue Yonder and NVIDIA are working together to fine-tune the Nemotron open source model for agent development
Blue Yonder, the supply chain AI company, today announced a model training factory built on NVIDIA Nemotron to accelerate the development of specialized AI agents for autonomous supply chains.
Model Training Factory, announced at ICON, Blue Yonder’s annual customer conference, is a repeatable system for fine-tuning and testing highly specialized supply chain models. Trained to consistently perform high-value tasks at the level of supply chain subject matter experts, these models are fine-tuned and built to perform complex, multi-step supply chain workflows in collaboration with human operators, and graded to ensure high-quality results. Accessed through agent AI, supply chain processes will ultimately be able to run autonomously, driving decisions across warehousing, supply and demand planning, transportation, merchandising, and network operations.
Blue Yonder and NVIDIA are working together to fine-tune the Nemotron open source model for agent development, building and deploying a system that combines NVIDIA’s Nemotron open source model and NeMo AI tools with Blue Yonder’s 40 years of supply chain decision-making, data, and operational expertise.
why is it important
Supply chain decision-making is highly complex and requires real-time analysis and coordination between globally distributed teams. You need extreme accuracy with extremely low latency across thousands of warehouses, lanes, and stores.
The next generation of AI assistants will help organizations analyze what’s happening in their supply chains faster. It enables more advanced and accurate AI and faster processing that is only possible with agent AI. Companies are moving from AI assistants to teams of specialized agents that can perceive, reason, use tools, and act alongside human operators at machine speed. The economics of doing so at scale are changing rapidly. The cost of running large-scale frontier models in production continues to rise as coding agents rapidly increase the demand for inference.
Model factories address these issues with a hybrid approach. Use Frontier models when you need a wide range of models, or custom supply chain models trained to work alongside them, to provide the accuracy and speed your individual workflows demand at a fraction of the cost.
“Supply chain has always been the realm of AI, and our research into how agent models work in real-world warehousing and planning decisions is exactly why we know where frontier models hit walls,” said Duncan Angove, CEO of Blue Yonder. “Together with NVIDIA, we are building owned intelligence, not rental intelligence: a supply chain model trained on the workflow, telemetry, and decision logic that actually runs warehouses and planning systems. This is not a one-off, fine-tuned model. This is a factory, producing specialized agents at the speed, accuracy, and cost that autonomous supply chains demand.”
Inside the model factory
Blue Yonder uses NVIDIA’s agent AI stack to build its Model Training Factory. We build on the Nemotron open model and use the NVIDIA NeMo Agent Toolkit to build, evaluate, and tune agents. Nemotron’s model size family allows Blue Yonder to match model size to the job, from compact models tailored for high-frequency decision-making in the warehouse to large models built for complex multi-step planning.
Each model is trained to be an expert at a specific task and provide specific outcomes for agent decisions, and adheres to rigorous evaluation criteria before deployment and as it is improved over time. Models are trained on synthetic data rather than customer data. Blue Yonder also uses NVIDIA AI Enterprise for its model training factory, combining microservices, frameworks, and libraries for AI development with advanced GPU orchestration and infrastructure management to deliver a fully supported, production-ready commercial software solution.
“The next stage of enterprise AI for supply chain requires specialized, affordable, domain-trained, accurate agents who can operate within the workflows that run a business,” said Azita Martin, vice president and general manager of retail and CPG at NVIDIA. “Blue Yonder is leveraging NVIDIA Nemotron, NVIDIA NeMo Agent Toolkit, and NVIDIA AI Enterprise to build a model training factory that uses its own supply chain data to fine-tune models, enabling us to build agent AI systems for some of the world’s largest and most complex supply chains.”
First proof point in the warehouse
Blue Yonder plans to deploy its first model for warehouse management workflows that include WMS allocation shortfalls, inventory exceptions, deadline urgency, and inventory across yards and receiving trailers. These are high-frequency warehouse decisions whose speed and accuracy directly impact on-time performance, stock shortages, and order cycle times. Subsequent models will expand into the broader Blue Yonder solution portfolio.
In a warehouse, shifts can quickly fall apart. The plans you set in the morning regularly change due to truck delays, equipment failures, and changing priorities, forcing you to constantly reallocate in a race against time. Expert agents can weigh hundreds of tradeoffs in seconds, where humans would typically consider fewer tradeoffs, and do so at a cost low enough to run continuously in every warehouse, every day.
reproducible benefits
Model Factory transforms operational expertise into reusable AI training signals and encodes that intelligence across domains in a repeatable way that can be extended across the supply chain. Blue Yonder’s advantage lies in the loop itself: workflow, decision logic, telemetry, subject matter experts, assessments, managed retraining, etc. that cannot be easily replicated by competitors. The first models are expected to enter customer production later this year through Blue Yonder Cognitive Solutions.
NVIDIA is helping Blue Yonder build the foundation for a new generation of supply chain AI. That means it’s open at the model layer, specialized at the workflow layer, and built to scale across enterprises that move goods around the world.
