Deutsche Telekom uses Celonis process intelligence and AI to save €5 million

Vendor messaging is often just that: messaging. Sounds good, but falls down in practice. What's more valuable is when customers speak openly about their projects in ways that directly corroborate - or contradict - what vendors are claiming. This week at Celosphere in Munich, Celonis took to the stage to argue that AI deployments will largely fail without process intelligence providing operational context. And throughout the week this claim was supported by multiple customer sessions at the conference.

Deutsche Telekom's case stood out as a particularly interesting example. Sebastian Dahs, VP of Service Process Improvement at the German telecommunications giant, shared how the company is using Celonis' process intelligence platform, combined with machine learning and AI, to predict which customers are heading toward problematic experiences - and intervening proactively before they churn. The result: approximately €5 million saved over six months, with critical customer cases reduced by 10,000.

The project ...