Build sovereign AI with Azure Local and Foundry Local | BRK442
Anders Zwartjes and Kushmeen Kambow present live demonstrations of local AI capabilities for datacenters, factories, branches, and sovereign environments, including Azure-consistent management.
Running AI in customer-controlled environments
The session explores how Foundry Local on Azure Local brings together several parts of an AI workload:
- Model deployment and scalable inference within customer-controlled infrastructure.
- Model evaluation without evaluation data leaving the environment.
- Agents and retrieval as components of locally operated AI applications.
- Azure-consistent management for these deployments.
Deployment and operation patterns
The demonstrations cover sovereign AI and disconnected operations, alongside multi-GPU inference. These patterns address running and scaling AI workloads in environments where control over infrastructure, data location, and connectivity matters.
This is an intermediate-level Cloud & AI breakout session from Microsoft Ignite 2026, identified as BRK442.