Regulated estates fail quietly. An NSG rule changes. A tag baseline disappears. A PIM window stays open past the ticket that justified it. By the time an auditor asks for evidence, the control plane is recounting history instead of enforcing it.
Ne Plus Ultra treats that as an operations product problem, not a quarterly workshop. The Zero-Drift Standard watches Azure Landing Zone resources continuously. When shape diverges from baseline, remediation starts automatically — policy, tags, network controls — and the estate returns to the measured state described in The Baseline Before the Model.
What ZDS must show
Telemetry is useless if it is a mock strip. Operators need resource identity, landing-zone context, open findings, and the last auto-remediation timestamp. The control plane surface at /operations/drift is built for that: hourly sweeps, estate health, and a one-click remediation path when the automatic loop needs a governor push.
JIT is a measured window
Just-in-time privileged access only works if revoke is immediate. A four-hour maximum window is the commercial cap. The hard SLA is revoke under thirty seconds — no staged delay, no waiting for a change window. If drop latency exceeds the SLA, the control plane records a miss. That is how standing privilege stays liquidated instead of renamed.
Why this belongs on the public site
Buyers do not need another diagram of zero trust. They need a path from stack composition to intake, and an operations plane that keeps the estate honest after the first deploy. That path starts in the configurator, continues through estate intake, and is governed in the control plane described under architecture.
If your landing zone still requires a human to notice drift, or a privileged session can outlive the ticket, the estate is already writing next quarter's finding.