Mondelēz picks Celonis to scale AI across operations

Mondelēz International has selected Celonis as a strategic partner for its enterprise AI programme as it moves towards what it calls an autonomous business.

The food group plans to implement the Celonis Context Model to combine process intelligence with predictive scenario planning across its operations. The work builds on a collaboration that began in late 2024.

The partnership sits at the centre of a broader effort to apply AI to large-scale business processes rather than isolated pilot projects. Mondelēz's process landscape spans manufacturing, shipment tracking, supplier payments and other functions across a business with annual revenue of USD $39 billion, about AUD $56 billion, and 91,000 employees serving more than 150 countries.

Executives said the focus is on building what they describe as self-healing processes that can identify issues early and adapt before they disrupt operations. That starts with connecting process data across the business rather than relying on large language models without operational context.

"Our ambition is to become an AI-driven business, and Celonis helps us fast-track that journey," said Rossana Rizzotto, Head of the Office of the CIDO at Mondelēz International.

"It enables us to position our ERP modernisation as a value-driven transformation and provides the digital foundation needed for AI-enabled operations.

"Getting iconic snacks like Oreo, Cadbury or Milka into the hands of billions of customers in more than 150 countries takes some heavy lifting behind the scenes. At our scale, hidden process problems can directly affect how we make our chocolate, track retailer shipments or pay suppliers.

"That's why we are racing towards becoming an autonomous, AI-driven business. But as any technology leader will tell you, there is a big difference between launching a single pilot and scaling artificial intelligence across business units and countries.

"At Mondelēz, we've seen first-hand that without business context, even the most advanced LLMs and agents risk making the wrong call. For us, the path to enterprise AI started with the core fabric of our business: our processes."

The partnership has already produced financial returns across seven end-to-end processes. Mondelēz reported USD $10 million in value generated and USD $6 million recovered in 90 days by addressing credit memo leakage.

One immediate example came from procurement. Celonis connected Mondelēz's top 20 systems of record, giving teams a consolidated view of supplier claims and payment issues.

"At a click of a button, we now see a full dashboard of which suppliers owe us what amount, so my team can follow up," said Miko Maciejewski, Vice President of MDS Procurement, Legal, CGA and Sustainability.

"And this is just the beginning," he said.

Maciejewski's team recovered USD $6 million within a quarter of the system going live. Mondelēz is also using the platform to create what it describes as a living digital twin of its operations, covering processes such as Source-to-Pay, Order-to-Cash and Hire-to-Retire.

AI groundwork

That process view is intended to support the next phase of AI deployment. Mondelēz is using the system to identify where AI agents might deliver the highest financial return, monitor the downstream impact of automated actions across systems and compare the cost of human workflows with agent-led ones.

The emphasis on return on investment reflects a wider shift among large companies from experimentation to measurable operating gains. For Mondelēz, the programme is also tied to ERP modernisation, with process data used to track whether technology changes are producing operational improvements.

Rizzotto said the model is changing how teams work by moving staff away from routine manual intervention and towards oversight of automated activity.

"Celonis gives us the visibility to fundamentally change the way we work," she said.

"It allows our teams to shift from doing manual work to overseeing AI agent performance, stepping in only to handle complex exceptions, so they can focus on strategic work."

Celonis said Mondelēz's use of the platform shows how large multinationals are trying to scale AI only after establishing clearer operational context. That means linking core systems, mapping actual workflows and measuring the effects of changes before increasing automation across regions and functions.

"Mondelēz International is setting a new standard for the digital-first enterprise," said Florian Schewior, General Manager for the DACH Region at Celonis.

"Celonis is proud to provide the critical operational context needed for it to confidently scale AI and automation across its global operations."