As a proof of concept moves into a governed modernisation project, what needs to be established for the compute, storage, repositories and access the work will use?

The correct answer is the one stating that compute and storage must be sized for the workload, repositories must be set up to hold generated artifacts, and access controls must be defined for who can use them. Moving from exploration to a governed project means the compute, storage, repositories for generated artifacts, and access controls all need to be deliberately provisioned for the scale and audience of the real work, not inherited informally from earlier experiments.

The answer saying nothing needs to change is incorrect because proof-of-concept resources are typically undersized and loosely controlled, making them unsuitable for governed, ongoing use.

The answer saying each developer should provision their own compute, storage, and repositories independently is incorrect because independent, untracked provisioning by each developer is exactly the kind of sprawl governance is meant to prevent.

The answer saying provisioning can be deferred until the project is closer to migration cutover is incorrect because waiting until near cutover leaves the team building on infrastructure that was never sized or secured for the workload.

For the exam, treat provisioning as an early, deliberate step in standing up a governed project, not an afterthought.