MODaaS, token economics & agentic AI - Building scalable AI-native operations

Sessions in rooms

Presentation: Running AI like a system: From models to managed services 
This session explores how operators are managing AI as a production system, using modular architectures, and governed deployment models like MODaaS. 

• How telcos decide what runs as a service vs embedded AI 
• How to manage multi-model environments across vendors  
• What breaks first when AI hits production?  
• What a real MODaaS operating model looks like  

Panel: The agentic telco: Are operating models ready?
AI agents promise a step-change in automation, enabling systems that can reason, decide, and act across complex environments. But scaling agentic AI raises challenges around governance, cost, interoperability, and control. This panel explores how far CSPs have really progressed - and what still needs to change. 

• Where has agentic AI failed in production – and why? 
• How to govern and control AI agent adoption 
• Whether current operating models are fit for agentic AI 
• How CSPs manage cost, risk, and accountability 

Case study: Taming AI spend: Token economics learnings from the frontline
More details coming soon.