Clarification Required on AI Unit Consumption, Allocation, and Token Usage for Worker Concierge V2?

Summary:

Requesting clarification on AI Unit consumption, allocation scope, and token usage calculations for a Worker Concierge V2 workflow developed using Oracle AI Agent Studio and deployed within Oracle Fusion HCM.

Content (please ensure you mask any confidential information):

We have developed and deployed a Worker Concierge V2 workflow using Oracle AI Agent Studio within our Oracle Fusion HCM environment. We understand that Oracle Fusion HCM SaaS includes 20,000 non-pooled AI Units, and we would like clarification on how these AI Units are consumed and managed.

Specifically, we would appreciate guidance on the following areas:

1. Whether interactions with the published Worker Concierge V2 workflow consume AI Units from the included allocation.