2.2.3. Deployment and Provisioning

Deployment decisions affect where your AI workloads run and how they scale. Cloud-hosted services offer simplicity and automatic scaling, while containerized deployments provide control for edge scenarios, air-gapped environments, or strict data residency requirements.

Deployment patterns tested on the exam:

• Cloud-hosted: Standard deployment with automatic scaling and updates
• Container deployment: For edge devices, on-premises, or disconnected scenarios
• Commitment tier: Pre-purchase capacity for predictable workloads at discounted rates

Choose containers when data cannot leave your premises; choose cloud when you need automatic scaling and managed updates.

Create Azure AI Services Resources