Governing AI agents in regulated PDLC
AI-agent governance is not a policy appendix. It changes the production lifecycle, the control model, and the meaning of readiness.
Request draftField notes
Short frameworks for the AI production edge.
Draftable essays and advisory memos on the questions that emerge when AI agents, release governance, controls, and executive accountability meet.
The independent go/no-go layer
In mature environments, governance must challenge both sides: whether a control is justified and whether delivery is genuinely ready.
Request draftWhen controls create second-order risk
Controls can reduce risk, move risk, or create risk. The difference matters before the decision gate, not after production.
Request draftAgenticOps needs a decision spine
Autonomous delivery roles only scale when ownership, observability, exception paths, and decision rights are explicit.
Request draftFrom process owner to change architect
The advisory value is not owning a process. It is seeing where the system cannot yet make a defensible decision.
Request draftPersonal context as AI governance infrastructure
Knowledge systems, agent skills, and structured context change what a human expert can perceive, remember, and reuse.
Request draft