Set up AI solutions in Foundry
AI-103 · Lesson 2 of 15 · 10:26
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AI-103 exam (Azure AI Apps and Agents Developer Associate). In this lesson: Foundry resources and projects (governance and development boundaries, the default project, production topology), deployment options (serverless API, managed compute, Foundry Local) and deployment types (Global Standard, Data Zone Standard, Standard, provisioned, Batch, Developer), configuring deployments (name, version, capacity, guardrail, version upgrade policies, model lifecycle, Azure Policy), basic vs standard agent setup, and CI/CD with Bicep or Terraform, azd, OpenID Connect pipelines and an evaluation step. Includes three original practice questions.
Preview the topics
- A European company requires that prompts are only processed inside the EU. It still wants new models and a generous quota. Which deployment type fits best?
- A production deployment must never change model versions on its own: the team upgrades manually, after testing. Which version upgrade policy should they set?
- A GitHub workflow must provision and deploy an agent’s infrastructure and code, and the security team forbids storing any client secret. What should you set up?