Business-wide / 14 days / £2,000
We may tell you not to build it.
Vendor-independent AI engineers map the business, quantify where work and money get stuck, and rank what is worth changing first. We call it the AI Constraint Audit, because the question it answers is which constraint is worth changing first.
The fit call is free. The fourteen-day audit is a paid £2,000 engagement.
Constraints. Opportunities. Economics. Priority.
The category, in plain words
An AI readiness assessment is a paid, time-boxed review that answers one question before any money is committed to a build: which operating constraint is worth changing, and whether AI is a viable way to change it. Every opportunity ends on one of four answers, build, buy, connect or stop, and the first AI decision is not which tool to buy.
Most AI readiness assessments sold in the UK are a questionnaire that returns a percentage and a sales call. This one produces four documents, an investment order and a fixed invoice, and the engineers who would build whatever comes next are the ones who run it.
What £2,000 buys
Constraint map, AI opportunity map, economic cases, ranked build proposal. An AI readiness assessment that ends in four working documents tied to evidence from across your operation, not a deck of generic AI ideas.
Swipe through the four deliverables
A business-wide view of where capacity, cash and decisions are held up. Work, waiting, hand-offs and rework are separated instead of guessed.
What a page of it looks like. A page of it is a table of process steps. Each row splits elapsed time four ways into work, waiting, hand-off and rework, names the evidence it came from, and carries a monthly total in hours.
Every viable opportunity, the dependencies underneath it and the points where a person must remain in control.
What a page of it looks like. A page of it is one opportunity per block: the step it changes, the systems and data it depends on, the decision a person must keep, and the condition that would make it fail.
Upside, cost, assumptions and risk for each opportunity worth considering, using evidence from your operation.
What a page of it looks like. A page of it is one case per opportunity: upside, cost to build, cost to run, and the assumptions behind each. Every assumption is labelled as ours or yours, so you can argue with it.
A prioritised investment order and custom proposal for the highest-value justified scope. You own the work whether AGMM builds anything afterwards or not.
What a page of it looks like. A page of it is an ordered list with the reason for the order, then a scope for the top item: what would be built, what it would cost, what it would not do, and what your team would have to supply.
AGMM publishes no redacted sample page of these four documents yet. The descriptions above are written from the real templates. The nearest inspectable thing is a whole issued job from another product: £98,214.72 across 63 measured lines.
YOU KEEP THE WORK WHETHER AGMM BUILDS ANYTHING AFTERWARDS OR NOT.
The fourteen-day sequence
£2,000, fixed, published, and the same for every business that buys it. It covers fourteen calendar days and three sixty-minute working sessions with the engineers who would scope any build. Your team spends three hours in the room.
Find waiting, rework and operational risk across departments, then gather the evidence that shows what actually happens.
Test cost, upside, dependencies and failure conditions with the people inside the work.
Review the maps and economic cases, rank what deserves capital and define the justified custom build.
Fit, stated plainly
Established UK businesses turning over roughly £1m to £50m, where the owner or the operations lead sits in all three sessions. Below £1m there is rarely enough repeated process to map. Above £50m there is usually an internal team who should own this rather than buy it.
Who attends matters more than the turnover. If the people in the room cannot change how the process runs, the assessment produces a document nobody can act on.
Swipe through the three warning signs
Skilled people spend valuable hours retyping, chasing, checking and rebuilding information across departments. Nobody has priced the total drag.
Different teams can each name a possible use for AI. Leadership needs one economic view of what deserves capital first.
The answer may be one connection, a process redesign or no new tool at all. Another subscription should not be the default.
The honest half
Outside the fee
Answer these before you book
No form, no score, no email address. Count the ones you can answer with a straight yes. They are about evidence and authority rather than technology, because those two decide whether an assessment produces anything usable.
Can you name the one process that cost the most time this month?
Can somebody say how many hours that process takes, without guessing?
Does a record of it exist anywhere other than in people’s heads?
Is there one person who can decide to change how it runs?
Would the people doing the work talk to us openly about it?
Is the data it depends on stored somewhere a system can read?
If the answer were “do not build anything”, could you accept that?
Could the business fund a build in the next twelve months if the case holds?
Six or more yes. The AI readiness assessment will produce an answer your team can act on. Three or fewer. Book the free fit call anyway and we will tell you what to fix first.
Day 15 onwards
On day 14 you hold four documents and an investment order. Four routes are open and three do not involve AGMM: commission the top-ranked build from us, hand the documents to another supplier, give them to a provider you already use, or fund nothing yet.
The AI readiness assessment is finished before any of that is decided, and no part of the work is withheld to make one route more likely than another.
The awkward question
Because a free assessment is still paid for, by the build it recommends. The £2,000 fee is not credited against a build, so there is no incentive to recommend one. That is the whole answer, and it is why the price sits on this page rather than behind a call.
It also sets the bar for our own work. A document somebody paid £2,000 for has to survive a finance director arguing with it.
Proof before promise
The assessment does not borrow another business's result and present it as yours. Every opportunity is tested against your own evidence before it earns a place in the order.
What we can show you today is a system AGMM built doing real commercial work: an issued construction quote of £98,214.72 across 63 measured lines, published line by line so you can check the arithmetic yourself. It is not an assessment deliverable and we will not pretend it is.
Inspect the £98,214.72 jobFourteen days. Three working sessions. Four client-owned documents.
Fit before invoice
Bring the process you suspect is costing the most. In 30 minutes we test whether it is specific enough to assess, whether the economics justify the engagement and what evidence we would need from your team.
Book the free fit callBefore you book
No. The AI Constraint Audit is a fixed £2,000 engagement. Only the initial 30-minute fit call is free.
This one is not. Assessments given away are paid for by the build they recommend. AGMM charges £2,000 and does not credit it against a build, so the fee does not depend on the answer. Only the 30-minute fit call is free.
In practice they are the same purchase under two names. AI audit also means inspecting AI systems already running, and statutory audit work done with AI, so the word is crowded. This engagement is the decision before a build: £2,000, 14 days, four documents you keep.
We test whether the workflow is specific enough to audit, whether the commercial stakes can justify the work and what evidence your team can provide. If the audit is not the right next step, we will say so.
Fourteen days, three sixty-minute working sessions with the forward deployed AI engineers who would scope any build, and four client-owned outputs: the constraint map, AI opportunity map, economic cases and ranked next step.
You can ask AGMM to implement the recommendation, give the work to somebody else, use an existing provider or stop. The audit is complete before any implementation decision is made.
Established UK businesses, generally £1m+ in annual revenue, with enough team, process and operational complexity for integration to matter.
That is a valid result. The purpose is to avoid paying for the wrong build, not to force AI into a process that needs a simpler fix.