The Celonis Context Model (CCM), announced on 12 May, is designed as a persistent operational layer that tracks how enterprise processes actually execute: order flows, procurement cycles, production sequences. The argument is that AI agents operating without this context produce answers that look correct in isolation but fail when they meet the complexity of real enterprise operations.
A day earlier, Celonis and AWS announced an expanded partnership built around zero-copy data access. Enterprises running the Celonis Process Intelligence Platform can now query data directly from Amazon S3 via an Iceberg REST Catalog integration, eliminating the pipeline step that previously required extracting and moving data before AI could act on it. BMW Group, which is already using the integration, said it "significantly reduces unnecessary data movement while maintaining control over our process data."
It's time to move past the era of AI science projects to Enterprise AI that delivers meaningful business outcomes.
The Ikigai Labs acquisition, announced alongside the CCM launch, adds forecasting and decision-intelligence capabilities derived from MIT research. The deal is expected to close imminently; MIT will become a shareholder in Celonis and the company will hold exclusive patent rights to the underlying technology.
AI is only as good as the context it has. Every organization needs to give its Enterprise AI a holistic, living model of how a business truly operates.
Customers don't want to spend months migrating data before they can see value from AI—they want to put their data to work where it already lives.
The three announcements arrive in the middle of a broader competition among enterprise software vendors to establish themselves as the operational layer that AI agents depend on. Celonis's specific argument is that the problem is not model capability but context: AI needs a model of how the business works, not just access to its data, to reason correctly inside it.
