Question 1

What does the AI Adoption Curriculum Builder actually do?

Accepted Answer

It backward-designs an AI-adoption curriculum and grades whether the plan holds together. You enter your target outcomes — each mapped to a role, a sequence step, an hours cost, and a Kirkpatrick measure (reaction / learning / behavior / results) — plus a weekly time budget and a rollout window. It scores four weighted dimensions into a 0–100 coherence score and returns COHERENT / LOOSE / FLAWED, then names the dimension to fix first. You get a runnable engine, a curriculum workbook, two playbooks, and a worked sample.

Question 2

How is the coherence score calculated?

Accepted Answer

Four weighted dimensions that sum to 100: role fit (25) — every outcome sits on a role that does that work; sequence & spacing (25) — foundational before advanced, prerequisites first, spread over weeks; outcome-measure coverage (30) — each outcome carries a Kirkpatrick measure; and time-budget realism (20) — total hours fit the weekly budget across the window. Each dimension is scored 0–5, weighted, and rounded to a 0–100 total: COHERENT at 75+, LOOSE 50–74, FLAWED below 50.

Question 3

Why did a plan scoring 94 still come back FLAWED?

Accepted Answer

Because of the hard gate. Any single target outcome with no valid Kirkpatrick measure forces the whole plan to FLAWED, regardless of the weighted total. The shipped sample scores 94 and is still FLAWED because one outcome (O5) has no measure attached. The score tells you how good the plan is on average; the gate tells you whether you can actually prove every outcome happened. An outcome you can't evidence is one you're only hoping for.

Question 4

How is the gate different from the measure-coverage dimension?

Accepted Answer

Coverage is continuous — the share of outcomes that carry a measure, scored 0–5 and worth up to 30 points. The gate is binary on the worst single outcome. A plan can score 4 of 5 on coverage (most outcomes measured) and still trip the gate because one outcome is unmeasured. Coverage tells you how measurable the plan is overall; the gate refuses to call it ready while even one outcome can't be evidenced.

Question 5

Does this score or rank the people on my team?

Accepted Answer

No. It grades the structure of a plan you enter — roles, sequence, measures, and time budget. It does not score, rank, or assess any individual person, and it must not be used for hiring, performance, or any other employment decision. It reads no systems and trains no one; it makes the plan honest, not the people. Educational; not legal, HR, or financial advice.

Question 6

What do I get, and what do I still need?

Accepted Answer

You get the coherence-grading engine, the curriculum workbook, a backward-design playbook, a rollout-and-sequencing playbook, and a sample plan with its worked verdict. It plans and grades; it is not an LMS or course content, so you deliver with your own training assets. It pairs with the AI Literacy & Workforce Training Kit (the program content your outcomes sequence toward), the AI-First SOP Engine (where workflow outcomes land), and the Claude Project Blueprint Library (per-function reusable setups).

#	Outcome	Role	Step	Hrs	Prereq	Measure
O1
O2
O3
O4
O5

Plan the AI rolloutbackward from the outcome.

A training plan can look full and still teach nothing you can prove.

Enter a plan and watch the coherence verdict recompute live.

The shipped example scores 94 and is still FLAWED — because one outcome had no measure.

Four weighted dimensions, summing to 100, with one hard gate.

It grades a plan. It does not score, rank, or assess people.

Anyone responsible for getting a team to actually use AI.

Plan here; deliver and sustain with the assets each outcome lands on.

Direct answers about what the Curriculum Builder grades — and what it doesn't.

Stop hoping adoption happens.
Plan it so you can prove it.