State Machine

An architectural model in which a business is defined as a set of possible states, the conditions that trigger transitions between them, and the deterministic logic that executes each transition — replacing human initiation and approval at each workflow step with encoded rules that run without human involvement.

The state machine, as applied to autonomous business design, is the structural model that makes human-independent operation architecturally precise rather than aspirationally described. In its computer science form, a state machine is an abstract model of computation in which a system exists in exactly one of a finite set of states at any given time, transitions between states in response to defined inputs, and produces defined outputs as a result of those transitions. Arco applies this model at the business level: every operational process is defined as a series of states, every trigger that changes the state is documented, and every action that executes in response to that trigger is encoded as deterministic logic. The business does not require a human to decide what to do next. The state machine already knows, because every possible next state has been defined before the first transaction was processed.

The practical consequence is the elimination of the coordination layer that human-centric businesses require to function. In a conventional business, the transition between operational steps is managed by people: a human reviews the output of step one and initiates step two, escalates to a manager when an exception occurs, and routes the output to the correct department through a combination of documented process and institutional knowledge. Each of these actions is a coordination cost. In a state machine governed business, each of these actions is encoded as a rule: if the output of step one meets the defined criteria, the system transitions to state two and executes the corresponding logic without human initiation. If it does not meet the criteria, the system transitions to an exception state and surfaces the condition to the Steward with the full context required for resolution.

This is the architectural expression of the Execution Layer and Judgment Layer separation. The Execution Layer owns every state transition that can be resolved by the defined logic. The Judgment Layer owns every transition that the architecture could not anticipate — the states the system was not designed to handle. The Intervention Threshold for each task tier defines the boundary between the two: what proportion of state transitions are expected to resolve autonomously, and what proportion are expected to surface to the Steward. At T1, Arco targets a 1:100 threshold — one Steward intervention per hundred autonomous state transitions.

First used: April 2026