Making the case for context-aware AI

What happened: Devavrat Shah, chief scientist of enterprise AI at Celonis, argues that enterprise AI agents need operational context to support decision-making within business processes, and recommends starting with a specific business decision and testing whether AI improves its hardest steps. Why it matters: Shah says agents without operational context risk overlooking constraints, making costly mistakes, or consuming excessive computing resources. He also says capturing the benefit requires investment in workforce education so employees can understand AI outputs.

Devavrat Shah, chief scientist of enterprise AI at Celonis, argues that enterprise AI agents need operational context to support decision-making within business processes, and recommends starting with a specific business decision and testing whether AI improves its hardest steps.

Shah says agents without operational context risk overlooking constraints, making costly mistakes, or consuming excessive computing resources. He also says capturing the benefit requires investment in workforce education so employees can understand AI outputs.