What would it take to put AI on the work that repeats?
Pick one piece of recurring work. Fourteen questions, about four minutes. You get a plain read on what a role around that work would need, and what you already have.
No email needed to see your result.
Seven conditions, two questions each.
A role does not need a perfect score on all seven. It needs enough on each. The assessment tells you which ones are already in place for your work, and which one is holding it back.
Value
Which business number this work moves, and who has committed to changing it.
Workflow
How well the steps, decisions and handoffs are understood, exceptions included.
Context
Whether the knowledge the work depends on sits somewhere a role could reach.
Systems
How work moves between systems today, and whether audited access has precedent.
Controls
Who reviews the output, and what a role would be allowed to decide alone.
Ownership
Who signs off, and how the people doing the work feel about changing it.
Measurement
Whether today's performance is known, and who checks it six months later.
Readiness belongs to the work.
Most readiness quizzes score the organization. That produces a verdict nobody can act on, because organizations do not get automated. Work does.
So this assessment starts where the buying decision starts: one piece of recurring work, named by the person who lives with it. A company that looks unready usually has one workflow that is ready, and finding that workflow is the whole job.
Scoring the organization
One verdict about a whole company. No work named, nothing to hand over, and no way to tell which team it applies to.
Scoring one piece of work
The work is named first. Seven conditions get scored against it, so the result points at one thing to settle rather than a programme.
Before you start
How long does the assessment take?
Fourteen questions about one piece of work, and about four minutes. You see your result without giving an email address.
Why does it assess one piece of work instead of the company?
Because readiness applies to work, not to organizations. A company can be described as ready and still have nothing specific to hand over, and a company that looks unready often has one workflow that is perfectly suited to a role.
What if our data is messy?
Everybody has data and nobody has clean data. A gap in context or systems tells you what the launch scope has to cover. It is not a reason to stop.
What happens to the answers?
Scoring happens in your browser, so nothing is submitted until you ask for the full report. If you do, the result is stored against your contact record so a conversation starts from what you already told us.
What is an AI role?
One piece of recurring work with a defined job, the systems it needs to finish it, and a person who stays in charge. Roles start behind human review and earn authority task type by task type.
Can I assess more than one piece of work?
Yes, and most people should. Run it again on the next candidate. Each completion is kept, so a second run does not erase the first.
A person in the room, same job.
An AI Workforce Review looks at where recurring work is costing the most time and margin, and maps the first one or two roles around it.
45 minutes. Nothing to prepare.