Skip to main content

Stewardship Model

The Arco operating model in which a single competent operator oversees an agentic stack, acting as architect and exception handler rather than executor.

Extended Definition​

The Stewardship Model is the defining differentiator of the Arco approach. Pure agentic deployment — where AI operates with no human oversight — is not yet appropriate across all task tiers. The Stewardship Model places one skilled operator in a supervisory role over the agentic stack.

The steward does not perform tasks. They design workflows, handle exceptions, and improve the system over time. This is categorically different from a traditional manager who coordinates human workers.

Key performance target: Mean Time to Intervention (MTTI) greater than 72 hours. If the system requires human input more frequently than once every three days, the architecture needs improvement, not more staff.

  • Autonomous Business — The Stewardship Model is the human architecture that makes the autonomous business operationally viable: it maintains human accountability and oversight without requiring human execution.
  • MTTI (Mean Time to Intervention) — MTTI is the primary performance target for the Stewardship Model: a Steward whose system runs more than 72 hours between required interventions is governing the system as intended.
  • Workforce Arbitrage — The Stewardship Model realises Workforce Arbitrage in practice: one Steward governing an agentic stack replaces many Process Workers, capturing the full human-to-compute cost differential.
  • Operational Arbitrage — The Stewardship Model is the operating architecture through which Operational Arbitrage is captured sustainably: it combines near-zero marginal execution costs with maintained human oversight.

Articles​

References​

  • Lexicon — canonical definition
  • Wiki — extended entry

Metadata​

First used: 2026-03-05
Pillar: How We Think

Part of the Arco Lexicon Ecosystem — maintained by Arco Venture Studio