Celonis users save USD $10 billion with AI operations

Celonis said customers using its Process Intelligence Platform have achieved combined savings of USD $10 billion by optimising operations and applying the software to enterprise AI programmes.

The company positions process intelligence as an operational layer spanning applications and data sources, giving AI systems the context needed for automation and decision-making.

Chief Product Officer Dan Brown said enterprise technology architectures are shifting as organisations experiment with AI tools, but companies still need a consistent intelligence layer beneath new interfaces.

"Celonis is that layer for AI-driven operations. We connect and transform data across the entire technology landscape into a digital twin to provide the real-time process and operational intelligence that AI agents need to deliver true value," Brown said.

Customer results

Celonis cited customer examples across telecoms, manufacturing, energy and distribution. Deutsche Telekom saved millions in revenue after using the platform to identify at-risk customers and intervene earlier in customer journeys, according to the company.

Fujitsu reduced excess inventory by 20% after buying teams received AI-driven recommendations generated from operational data. Mercedes-Benz improved on-time delivery and shortened decision cycles across more than 30 production plants, the company said.

In energy, Uniper recorded double-digit millions in savings across 27 processes by orchestrating an AI maintenance agent. Vinmar increased operational productivity by 20% within a USD $3 billion business unit after orchestrating multiple AI agents, according to Celonis.

Celonis linked these outcomes to an approach it calls composable enterprise AI, which uses modular components rather than monolithic systems. It argues that process-level visibility and governance become more important as organisations add AI tools and automation to daily workflows.

Platform features

Celonis said the platform provides "horizontal operational context" across a company's systems. Brown said it helps identify where AI delivers the highest return on investment, sets guardrails for safety and compliance, and reduces dependence on legacy systems.

The platform includes a "Multimodal Digital Twin" that represents operations by combining process data with desktop actions and unstructured or semi-structured information. It can also incorporate data from outside an organisation, Celonis said.

Celonis also highlighted "zero-copy" bi-directional integrations with data platforms such as Databricks and Microsoft Fabric. These connections allow data to flow into the process intelligence layer and insights to flow back into data lakes and analytics environments.

Other components include an Annotation Builder, which Celonis describes as a low-code tool for enriching data with AI-generated information, and a Model Context Protocol server that exposes Celonis tools and process intelligence to third-party AI agents. Celonis said this enables grounded reasoning and automation using process context.

A further module, the Celonis Orchestration Engine, manages processes and coordinates people, AI and existing automation. Celonis said orchestration becomes more important as organisations deploy multiple agents and bots across overlapping workflows.

Growing roster

Celonis named additional companies it said are advancing enterprise AI projects on its platform, including Akzo Nobel, Alps Alpine, Daimler Truck, Karl Storz, Latam Airlines Group, Mann + Hummel, Mondelez Europe Procurement, Telia and Toyota Tsusho.

It also said it is expanding in Asia-Pacific, citing customers including True Alliance, Optus, Queensland Health and Zespri.

Brown described the platform as a foundation for enterprise AI efforts that span functions and systems.

"Momentum for the AI-driven, composable enterprise continues to accelerate," Brown said.