Building the Agent Fleet: Operating Model, Governance & Team Design

What would a global CRE department look like if it were designed from first principles for an agentic AI world. Should every repeatable CRE task be assigned to an AI agent as the primary executor?

What agents are you running, how did you design your agent architecture, what autonomy have you granted them, how did the team restructure and what governance framework made the difference between a pilot and Agentic-AI trustworthy enough to run at scale?