GoKawiilCelonis executive outlines design principles for its AI context layer
In an interview, Celonis's Devavrat explained that the company's 'context model' sits between workflow and system-of-record tools below and business users or AI agents above. He said the model is built to abstract underlying processes, deploy guardrailed agents for specific tasks, and be 'agent first' so other agentic systems can interface with it directly. He added that existing workflow data does not need to be complete or clean for the system to connect to and interpret it.
GoKawiil's interpretation of the reporting above, not reported fact.
The comments suggest Celonis is positioning its context layer as infrastructure that can sit atop messy, pre-existing enterprise systems rather than requiring costly data cleanup first, which could lower adoption barriers for AI agents in business workflows. Framing the layer as 'agent first' also indicates Celonis anticipates a future where multiple AI agents, not just human users, interact directly with enterprise process data.
- Celonis describes a three-layer architecture: system-of-record data, a context model, and agents or human users on top.
- The context model is designed to abstract processes, deploy guardrailed agents, and interoperate with other agentic systems.
- Celonis claims existing workflow data doesn't need to be complete or clean for its system to make use of it.
Source: technologyreview.com — Mit Technology Review Insights, 2026-10-22
Published there as: “Making the case for context-aware AI”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.