Fixed scope. Fixed fee.
We agree the scope and fee before the work starts.
-
The decision usually arrives before the certainty.
A vendor has made a proposal. A pilot wants more budget. A team wants to build. Someone wants to hire. The board wants to know what happens next.
At that point, the question is no longer whether AI matters. The question is whether this move deserves a yes.
An AI Decision Review is an independent review of the decision in front of you. We look at the business case, the economics, the operating reality, the data, the ownership, the risks, the technology and the assumptions underneath the proposal.
Then we tell you what we think. Proceed. Change it. Pause. Or stop. And why.
We do not sell the software. We do not implement the platform. We do not receive a commission if the project goes ahead. That is precisely the point.
What we review
Not whether the technology is impressive. Whether the decision makes sense for your organisation.
Value.
What changes if this works, and is that change worth paying for?
Cost.
What will this actually require beyond the number in the proposal?
Reality.
Do the process, data, people and systems support what is being promised?
Ownership.
Who is accountable when the demo becomes an operating system?
Risk.
What can go wrong, what matters if it does, and who carries that exposure?
Alternatives.
Is this the right move, or simply the first solution that reached the table?
What you receive
A written decision memo.
A clear recommendation: proceed, change, pause or stop. No maturity score for the sake of having a score. No generic benchmark. A position on the actual decision.
The case for and against.
Value, cost, dependencies, risks and assumptions laid out in plain language. Including the uncomfortable parts.
The conditions for a yes.
What has to be true before money, people or reputation should be committed. If those conditions are not there yet, we say so.
The next 90 days.
If the answer is yes, what happens next. If the answer is not yet, what needs to change first. If the answer is no, what you stop spending time on.
A 90-minute leadership debrief.
In the room with the people who have to decide. Questions answered. Assumptions challenged. Decision made.
This is not an AI audit.
We are not trying to give the organisation a score. And we are not trying to find twenty-seven AI opportunities so there is enough material for a deck.
The scope is simpler. There is a decision in front of you. We make it easier to make the right one.
When to call us
Before signing the vendor contract.
Before approving the pilot.
Before turning the pilot into a rollout.
Before hiring the AI team.
Before committing the budget.
Before the board asks why this was approved in the first place.
Sometimes the answer is that the decision is too early.
That can mean the organisation needs a broader AI strategy first. It can mean the economics do not work. It can mean the data is not there. It can mean the vendor is solving the wrong problem. Or it can simply mean that there is a better place to start.
We will say that too. In writing.
What happens after the review?
A Decision Review does not automatically turn into another engagement. Sometimes the answer is enough. When more work is needed, the next step depends on the decision.
-
No clear AI direction yet?
Strategy & roadmap. -
The direction is clear, but nobody owns it?
Fractional Chief AI Officer. -
A working pilot needs to become operational?
Pilot to production.
The next engagement should follow from the decision. Not the other way around.
You do not need another opinion on AI. You need to know what deserves a yes.
-
A roadmap is not a list of every AI use case in the building. It is the small set of bets worth defending in front of a board.
An AI strategy is not a long list of ideas. It is a shorter list of decisions: what to do, what not to do, what comes first, and what has to be true before the next step.
We work with you to write that document. We start from where the organisation actually is, we sit with the people who own the P&L, and we narrow. The output is short by design. Three to five strategic bets, sequenced over eighteen months, with named owners, named decision points, and the budget envelope each one sits inside.
It is the document the CEO can hand to the board, and the document the board can hand to the auditor.
What you receive
A written AI strategy.
Three to five bets. Why these, why now, why not the others.
An 18-month roadmap.
Sequenced quarters with owners, budgets and decision gates.
Governance scaffolding.
Approval routes, escalation paths, kill criteria.
A board-ready narrative.
One deck. One memo. Both defensible under questioning.
Roadmaps without owners are wishlists. We will not deliver one without naming the people accountable for each move.
-
AI rarely fits one department. That is why it often ends up owned by no one.
AI touches too many functions to float between them. We take the seat until the organisation knows what the permanent role should look like.
In the seat we set direction, choose priorities, structure the portfolio, prepare decisions, challenge vendors, set governance and bring the right questions to the table. The role is not there to make AI look organised. It is there to make the organisation decide better.
Some organisations need the function before they know the final job. Leaving the agenda loose is risky. Hiring too early can be just as risky. Fractional gives the work a place while the permanent shape becomes clear.
What you receive
Agenda ownership.
One named seat for AI in leadership meetings.
Portfolio cadence.
Monthly review of moves, owners and gating questions.
Governance frame.
How decisions are made, escalated and recorded.
Hand-over plan.
A profile and runway for the permanent CAIO.
AI needs ownership before it needs an organigram.
-
A pilot proves potential. Production proves whether the organisation can use it.
A demo proves that something can work once. Production asks whether it can work repeatedly, with real users, real data and real consequences.
After the demo, the data gets messier. The users get less forgiving. The edge cases appear. The output needs a quality standard. The organisation needs an owner. That is where many pilots stop moving, even when the demo was good.
We clarify who owns the output, what quality means, who reviews what, what happens when it is wrong, what gets reported, when to scale and when to stop. The point is not to keep the pilot alive. The point is to make the next decision unavoidable.
What you receive
Production readiness review.
What is missing between pilot and live.
Quality standard.
How good the output has to be, and who decides.
Ownership map.
One owner, one reviewer, one escalation path.
Stop criteria.
The conditions under which the pilot ends. On purpose.
Not every pilot should go to production. The waste is not stopping. The waste is drifting.
-
Everyone has seen what AI can do. That is no longer the hard part.
For boards, executive teams, leadership offsites and conferences. No vendor slides. No AI theatre. A point of view.
The hard part is deciding what it should do here, in this organisation, with these risks and these people. Typical questions for the room: why do AI pilots fail after a successful demo? What should a board ask before approving AI budget? Who owns AI when every function is involved? How much governance is enough? When should management stop a pilot?
No vendor pitch. No generic AI introduction. No future-of-work sermon. No stage noise. The room already has enough noise.
What you receive
Boardroom session.
A point of view, not a vendor deck.
Tailored questions.
Calibrated to your sector, your decisions.
Pre-read & follow-up.
A short brief before. A short memo after.
Optional Q&A workshop.
Working session for the leadership team.
The point is not applause. The point is the next meeting.