Supply Chain AI Partnerships Drive Growth

Gartner recently gathered supply chain and operations executives from five industries for a Leaders in Action event focused on applying AI to drive sustainable growth in partnership with internal and external solution providers. A huge “thank you” to Willem Uijen, Unilever’s chief supply chain and operations officer, and team for hosting our community.

The event highlighted the importance of key supply chain partnerships. For Unilever, one of those relationships is with the McLaren Formula One (F1) Racing Team. Supply chain organizations can learn from the McLaren team in areas like operational agility. The fastest pit crews can service an F1 race car in 2-3 seconds.

Innovation is another area. During our session — held at McLaren’s Technology Centre in London’s suburbs — the Unilever team highlighted that McLaren advised them on how to improve some of their factories’ manufacturing flows.

McLaren also values its partnerships, such as its relationship with Lego Group. The following picture shows off this full Lego replica of a McLaren car. It’s built around a real McLaren chassis and comprised of more than four hundred thousand bricks.

The Unilever team showcased AI-based capabilities in a wide variety of supply chain contexts, with the deepest dives in customer operations and sourcing/procurement. They also highlighted areas where they evolved and streamlined their supply chain processes to get AI-ready before outsourcing them to an external business partner.

In the customer operations space, Unilever developed an AI platform called SKY that leverages several dozen machine learning routines to analyze real-time, store-level customer data. By connecting it to forecast and inventory information, Unilever enables enhanced replenishment planning, driving superior product availability for key customers.

One of the more sophisticated AI tools that Unilever built for sourcing across the four-tier palm oil supply chain is a model that incorporates both internal and external data (in some cases, unstructured) to project future commodity pricing and optimize purchase timing and sourcing based on a dynamic view of the marketplace.

Rounding out the discussion, Cliff Henson, Corporate VP for Microsoft Cloud’s supply chain, gave an inspiring and thought-provoking presentation. He started by highlighting his team’s progress in deploying ten AI assistants and agents across a span of supply chain functions, then laid out Microsoft’s plans to deploy more than one hundred(!) agents in 2026.

These agents’ scopes vary. Some will cover subsets of workers’ overall responsibilities, while others might be orchestrator agents that tap into multiple smaller ones to accomplish broader tasks. The largest-sized agent would require approximately six weeks of development time, and leveraging a standard agent-creation platform will enable smaller teams of software engineers and business experts to create their own agents with a much lighter product and project management footprint.

Other key takeaways from this Leaders in Action event include:

• Unilever and Microsoft’s heads of operations both own the IT function for their supply chains and Unilever’s leader also serves as corporate CIO. It’s an interesting trend that we’re starting to see among the more mature organizations in our global COO/CSCO community. Integrated ownership helps facilitate a more cohesive strategy for technology adoption and allows full ownership of both the business value and delivery of technologies like AI.
• Many of the supply chain executives in the room acknowledged that more-advanced forms of AI will both augment workers to enhance their performance and, in some cases, mean less workers needed in specific functions. For organizations continuing to grow top-line revenue, this may mean holding headcount flat or not backfilling attrition.
• If the scope of agents deployed becomes broad enough, such as in the long-term vision for Microsoft Cloud’s supply chain, eventually the agent platform may evolve into the primary user interface for some end users. This assumes that agents are both reading from and writing to underlying ERP and legacy systems.

We’re grateful for the inspiring two days spent together in London and look forward to continuing these discussions with the COO/CSCO community at our future events.

Stan Aronow
VP Distinguished Advisor
Gartner Supply Chain
Stan.Aronow@gartner.com