AI-300 Practice Question 16: Data Zone Deployment Type
Microsoft Machine Learning Operations Engineer Associate (AI-300) · Test 2 · question 6 of 10
A German insurer is deploying a chat model in Microsoft Foundry for an internal claims assistant. Legal requires that prompts and completions be processed only within the EU data zone. Traffic is spiky and unpredictable, finance has refused any capacity commitment, and the team wants the highest default quota still available under the residency constraint. Which deployment type should you choose?
Show answer and explanation
Answer:
Microsoft's deployment type table lists Data Zone Standard as processing "within data zone", billed pay-per-token, and best for "EU/US/APAC data zone compliance", and the requirement mapping says plainly: "Keep processing within a data zone → Data Zone Standard or Data Zone Provisioned". Between those two, only the Standard variant avoids a commitment, and data zone deployments "provide higher default quotas than geography-based deployment types while keeping data within the data zone" (option C).
- A is wrong because Global Standard routes traffic to any Azure region. The highest quota in the catalog is irrelevant when it breaks the one hard requirement legal set. - B is wrong because Data Zone Provisioned satisfies residency but is billed as reserved PTU, the capacity commitment finance rejected, and reserved capacity is a poor match for spiky, unpredictable traffic. - D is wrong because the Standard type confines processing to a single Azure geography and is documented for geography compliance and low volume. It is stricter than the requirement and carries lower default quota than the data zone type, so it gives up quota for nothing.
Reference: Understanding deployment types in Microsoft Foundry Models