The agents' questions compel student tutors to think and explain the materials in different ways, and watching the agent solve problems allows them to see their knowledge put into action.
Once these different functions (such as the probability distribution) are learned, the correct action to take is simply a matter of deciding which action maximizes the "expected utility" of the agent.
Once the VM is fully instantiated, the Workload Deployer-specific agent residing inside the VM initiates the action to configure the VM for the role it will play in this application deployment.