Free executive playbook.

One human, thirty agents — real, but earned. The ratio is an output you earn, not an input you pick. This playbook is the operating model that makes a high human-to-agent ratio safe: the four pillars that hold the fleet, the five forces that set the safe span, and the Span Audit you can run on Monday — so one human supervises the right number, not the most.

What's inside

  • The ratio, earned — why a high ratio fails without proportionate oversight (the 700:1 cautionary tale), and the shadow agents that live between identity sprawl and the safe span
  • The four pillars — Span of Supervision, the Human Role Shift, Orchestration Architecture, and the Control Plane
  • What bounds the span — the five forces that raise or lower the safe ratio, and why speed of control beats span of control
  • Regulation as the control plane's spec — SR 26-2, CFPB, UK SMCR/FCA, EU AI Act Article 14, and Vietnam, converging on one minimum control set
  • The Span Audit — a four-step loop to size, orchestrate, and govern your fleet

Five pages · free · PDF.

Want the weekly version? Subscribe to The AI Architect Letter · Read the latest insights.

Independent thought leadership · not affiliated with any current or past employer · compliant with Vietnam AI Law 134/2025 + PDPL.