Skip to content
Diagnostic design

The operating judgment behind the diagnostic.

This diagnostic is derived from an intake-to-decision operating system. The summary below shows where AI can assist, where human judgment must hold, and how each diagnostic dimension maps back to a workflow stage.

The design is organized around decision quality, capacity allocation, and delivery predictability as AI enters the workflow.

01Decision rights
AI prepares the work

Prepares, detects, routes, summarizes, and drafts option sets.

Humans hold the decision

Approve, fund, defer, narrow, accept risk, override, and commit capacity.

02The bounded, observable, reversible test

Any agentic AI placement must be bounded in scope, observable in effect, and reversible if wrong. Placements that fail any of the three do not get made.

03The executive decision moment

AI is deliberately absent at the executive decision moment. The decision right stays with an accountable person, and the rationale is recorded.

04Feedback into intake

Execution data returns to planning. Actual effort, cycle time, and handoff quality are compared against the original estimate. Lessons from delivery update the intake and planning system. Someone owns the decision to sunset or stop work that is no longer worth continuing.

05Dimensions and stages
StageDimension
01Demand Intake Discipline
02Problem Framing and Triage
03Capacity and Estimation Realism
04Cross-Functional Risk and Dependency Visibility
05Option Formulation and Tradeoff Clarity
06Decision Packet and Governance
07Decision Rights and the Executive Moment
08Execution Feedback and Learning Cadence