AI-300 Practice Question 39: Project Tag Cost Attribution
Microsoft Machine Learning Operations Engineer Associate (AI-300) · Test 4 · question 9 of 10
Four teams share one Microsoft Foundry resource, each working in its own Foundry project, and all use Azure OpenAI models that are sold by Azure. Finance needs each month's token spend split by team for chargeback, and the team accepts using the project-level cost attribution feature, which is currently in preview. The platform team does not want to create separate Foundry resources or add custom logging around every model call. What should they use?
Show answer and explanation
Answer:
Foundry supports *chargeback with project-level cost attribution*, which is currently in preview, so that *"FinOps teams and admins can allocate shared Foundry spend back to the business unit, team, or workload that incurred it."* In the documentation, *"every Foundry project is automatically tagged with a project tag on its underlying usage"*, so in Cost Management the team filters cost analysis by that tag and sees spend per project, without adding tags by hand. It is supported for models sold by Azure, including Azure OpenAI (option A).
- B is wrong because grouping by meter breaks costs down by model series and input or output usage, not by project or team. - C is wrong because splitting into separate resources is exactly the restructuring that the platform team wants to avoid, and the tag already provides the split. - D is wrong because custom counting duplicates data that billing already records, drifts from the invoice, and needs code around every model call.
Reference: Plan and Manage Costs - Microsoft Foundry