An honest price depends on your number. So we print the formula, not a shelf tag. You can work out the maths yourself before you sign anything.
No hidden ranges · No quote chase · Figures scoped per Diagnostic
Before anything is priced, the Diagnostic settles one number: the annual value of moving the metric that actually leaks. Not a guess, not a range hidden until the call. A figure we both say out loud and both agree on. Every price below is a plain fraction of that number, so you can do the maths yourself before you sign anything.
The read that sets the anchor. Priced by application at this stage, not by the hour. You leave with the diagnosis and the number whether or not a build follows.
A one-time, fixed price tied to the value the build releases, roughly a tenth of the annual number you agreed at the Diagnostic. Fixed scope, scored acceptance: the figure does not move once it is set.
A recurring base tied to the work it takes off your team, a fifth to a third of the monthly labour cost the systems replace. Flat maintenance plus the SLA tier you choose. It costs less than the hours it removes, or it is not worth running.
Every band is [scoped per diagnostic] until the Diagnostic sets your number. The formula above is how the band is reached, printed so nothing is decided on a call you can't check.
Sets the anchor number.
Priced off the anchor. One task, replaced.
Priced off the labour saved. Kept running.
The accounts, the code, the keys, all provisioned in your name. Stop any time and you keep everything that was built, still running, without us.
The value anchor is said plainly and agreed between us before we build a thing. No range hidden until the call, no figure sprung on you afterwards.
Scope is fixed in writing before you sign. A mid-build addition is a small, separate, priced thing, never a surprise on the invoice.
Third-party usage is billed straight to your accounts, in your name. No markup, no mystery line item, nothing padded.
Because an honest price depends on your number, and your number comes out of the Diagnostic. A build that moves a large annual figure is worth more than one that moves a small one, and pricing it as a flat menu item would either overcharge the small case or undercharge the large one. Instead the formula is printed here and the anchor is agreed with you out loud, so the price is something you can work out yourself. Not a figure we reveal on a call.
The scope is written before you sign: a not-doing list that says plainly what the build will and won't touch, revision caps, and a pre-signed change-request clause so a mid-build 'can it also…' becomes a small, separate, priced add-on rather than a surprise on the invoice. Acceptance is scored against an agreed bar by evals, not by vibes. The number set at the start is the number at the end.
You do. All of it. The accounts, the code, the automations, the data, the API keys: everything is provisioned in your name and handed over. The AI Partner retainer is how the system stays cared for, not how you stay locked in. If you ever stop, you keep everything that was built and it keeps running without us.
Every build carries a Day-30 review, baked into the price. Thirty days after a system goes live we sit down with the numbers: is it doing what the anchor said it would, what needs tuning, and whether the next module has earned its place. It is the checkpoint that keeps the partnership honest. The retainer grows by evidence, one module at a time, not by default.
Apply for a Diagnostic. We agree the number out loud, and every price above becomes something you can work out for yourself.