Blog

How Danone Transforms Dairy Planning Chaos into Clarity

Transform your complex business challenge into an optimized plan of action—powered by Gurobi’s world-leading solver technology.

Blog

How Danone Transforms Dairy Planning Chaos into Clarity

Transform your complex business challenge into an optimized plan of action—powered by Gurobi’s world-leading solver technology.

Blog

How Danone Transforms Dairy Planning Chaos into Clarity

Transform your complex business challenge into an optimized plan of action—powered by Gurobi’s world-leading solver technology.

author

Pamela Trevino

Director of Marketing

Pamela Trevino

Bio

author

Pamela Trevino

Director of Marketing

Pamela Trevino

Bio

Today’s businesses already face plenty of complex supply chain challenges—but imagine if your business was dependent on something completely out of your control, like how much milk a cow produces.

That’s exactly what Danone, the number one brand in milk and plant-based products in the world, faces every day as their planners try to answer key questions, such as, “How much and from which supplier should materials be sourced, in order to satisfy demand while minimizing costs?”

At the 2025 Gurobi Summit in Vienna, Shrya Gupta, Global AI Strategy Lead at Danone, shared how the company is leveraging optimization to successfully manage mid- and long-term planning despite fluctuating supply and demand.

Overcoming Uncertainty to Balance Supply and Demand

Milk demand is nonlinear, varying significantly based on promotion and consumption seasonality. And of course, cows cannot produce milk on demand, which adds even more variability. When milk demand exceeds internal milk collection, deficits may occur; and when internal collection exceeds demand, excess milk is produced.

Because of these uncertainties, milk planners are left with a complex equation—with over 100,000 constraints and 10,000 variables—that must be solved in order to satisfy demand and minimize costs.

Key considerations include:

  • Production constraints at each plant (19 plants across 5 countries)

  • Flexibility in the formula used for production

  • Milk production capacity

  • Seasonality of protein and fat contents

  • External partners that bring different transformation capacities

  • Supplier capacity to purchase or sell dairy ingredients

For milk planners, it could take days to weeks to consolidate a single scenario—with no guarantee of accuracy or optimality. In order to maintain competitive costs, flexibility, and adaptability to evolving business requirements, Danone developed Deep Blue, an in-house optimization solution powered by Gurobi.

Managing Complexity with Deep Blue

With the goal of minimizing sourcing costs while meeting demand for material at each factory, DeepBlue utilizes a constrained optimization method powered by Gurobi to create sourcing strategies, plan mid-term and long-term milk supply, and manage operations and farmer relationships.

The solution addresses key complexities in milk sourcing, such as demand fluctuations, fixed contracts, and the need to balance supply and demand.

Developing an in-house optimization solution (rather than using an off-the-shelf milk optimizer) offers several benefits to Danone, including:

  • Competitive costs, which are regularly reassessed

  • Internal knowledge of the Danone network and supply chain model

  • Flexibility to adapt to evolving business models

  • The know-how to replicate for other domains

And while this approach also came with several challenges—such as the need to standardize processes to make the most out of the tool at a minimum cost, and find the right balance between user-friendliness and complexity for the data science team—those challenges also presented important learnings.

For example, when Deep Blue first launched, it took new data scientists an average of three months to be fully onboarded to the project. But by keeping their documentation well updated, removing unnecessary complexities, and making sure all team members understand the core components, they’ve been able to reduce the learning curve to just three weeks.

Moving forward, Danone plans to expand the use of Deep Blue to their North American operations and explore further optimization opportunities, such as maximizing the use of raw materials in their production processes.

“GenAI or AGI may not be able to solve the problem of cows meeting the demand that Danone has for milk, but Deep Blue, powered by Gurobi, is able to do that and we are very proud of it,” shared Gupta.

Download the Complete Presentation

The success of Deep Blue demonstrates Danone’s commitment to leveraging advanced analytics, and the power of optimization to drive operational efficiency and meet the company’s milk sourcing needs.

Want to learn more about how Danone optimizes milk sourcing operations? Check out their complete presentation from the Gurobi Summit in Vienna here.



Start Solving with Gurobi

Try Gurobi on your own optimization models and see how it performs on real decision problems.

Start Solving with Gurobi

Try Gurobi on your own optimization models and see how it performs on real decision problems.

Start Solving with Gurobi

Try Gurobi on your own optimization models and see how it performs on real decision problems.