Plan & Manage AI Solutions
This domain covers the foundational skills every Azure AI engineer needs: selecting the right services, provisioning resources securely, monitoring health, managing costs, and implementing responsible AI practices. It represents 20–25% of the AI-102 exam.
Before you write a single line of AI code, you need to understand the Azure AI services landscape, know how to secure endpoints with keys and managed identity, deploy services in containers for edge scenarios, and set up diagnostic logging that enables production troubleshooting.
These challenges establish patterns you'll reuse throughout every subsequent domain. Master authentication, resource provisioning, and monitoring here — they appear in every other section of the exam.
What You'll Learn
- How to select the appropriate Azure AI service for a given requirement
- Create and manage Azure AI service resources (Portal, CLI, Bicep)
- Configure network security (private endpoints, VNets, firewalls)
- Implement key-based and identity-based authentication
- Deploy AI services in Docker containers
- Set up diagnostic logging and monitoring with Azure Monitor
- Manage costs with pricing tiers and budget alerts
- Apply responsible AI principles and content filtering
Skills Measured
- Select the appropriate Azure AI service for a requirement
- Plan and configure security for Azure AI services
- Create and manage Azure AI resources
- Configure diagnostic logging and monitoring
- Manage costs for Azure AI services
- Implement responsible AI practices
- Deploy AI services in containers
Challenges
| # | Title | Key Topics |
|---|---|---|
| 01 | Select and Provision Azure AI Services | Multi-service resource, single-service, SKU selection |
| 02 | AI Services Resource Management | CLI/Bicep provisioning, tagging, resource organization |
| 03 | Secure AI Endpoints with Keys & Identity | AzureKeyCredential, DefaultAzureCredential, Key Vault |
| 04 | Network Security for AI Services | Private endpoints, VNet integration, IP firewalls |
| 05 | Deploy AI in Containers | Docker containers, connected/disconnected mode, billing |
| 06 | Diagnostic Logging & Monitoring | Azure Monitor, Log Analytics, diagnostic settings |
| 07 | Cost Management & Optimization | Pricing tiers, budgets, commitment tier discounts |
| 08 | Health Checks & Alerting | Availability metrics, alert rules, action groups |
| 09 | Responsible AI Governance | Transparency notes, impact assessments, RAI dashboard |
| 10 | Content Safety & Filtering | Azure AI Content Safety, severity thresholds, blocklists |
Prerequisites
- Azure subscription with Contributor access
- Azure CLI installed and authenticated
- Basic Python or C# programming skills
- Familiarity with Azure Portal navigation