Every stage has a deliverable.
The plan below is an example, not a quote. What is fixed is the shape: the order of the work, one thing finished at each stage, and the scope agreed in writing before any work starts.
The shape of an engagement
- Free 60-minute diagnostic
- A discovery, credited against the build
- Fixed scope, fixed price
- Then the care plan, in one of two sizes
The hour is free because it is not the work.
On a call or on site, and it works far better with the people who actually do the work in the room rather than only the person who signs. There is no deck. We ask what the week looks like and both sides find out whether there is anything here worth building, before anybody spends money finding out.
It is a qualification call. No document comes out of it and nothing is owed either way.
The measuring comes later. An hour tells us where to look; what the work actually costs you is what the discovery is for, and no honest number comes out of a conversation. The discovery is priced, and what you pay for it is credited against the build.
No is a good outcome. If the honest answer is that nothing in your week is worth automating yet, you have saved yourself the discovery.
Book a diagnosticAn example six-week plan, stage by stage.
An illustration, not a quote — the scope is made up. Yours gets its own dates, in writing, before anything is signed.
- 01
The discovery — where the measuring happens
We interview the people in the roles you nominate. If you would rather not give up the time, a chart of job descriptions and day-to-day tasks gets us most of the way. Nothing gets proposed yet. A system designed off a description of the work is a system that meets the description instead of the work.
From you Time from the people in those roles, or your job descriptions and task lists, and read access to the systems the work lives in.
You get Every repeating task we found, with an hours-per-week figure against it. Priced, and credited against the build.
- 02
The map, and the build order
What we measured gets written up and ranked: what we would automate, what we would leave alone, and what is a broken process rather than a slow one. If the measurement does not match what we were told in the first hour, the plan gets corrected here.
From you An hour to argue with the ranking. You know the business; we have only been watching it.
You get A written, ranked map of where the hours go, including the noes, and an agreed order to build in.
- 03
The first system
The biggest item on the list gets built, into the systems you already run. It handles real work on a narrow slice — one job type, one branch, one queue — so that if it is wrong, it is wrong somewhere small.
From you The person who owns that process, for a couple of hours.
You get One system running on live work, with a person checking every output.
- 04
The rest of the shortlist
The same pattern, faster, now that we know your data, your naming and your exceptions. The slice on the first system widens as it earns it rather than on a date.
From you Exceptions, as they surface. They will surface, and that is the useful part of this stage.
You get The shortlist built. None of it trusted yet.
- 05
Your team tries to break it
Everything runs on live work while we sit next to the people using it. This is the stage that finds the customer whose name has a slash in it, the job type nobody mentioned, and the Friday afternoon when three things happen at once.
From you Your team using it for real, and telling us plainly what annoyed them.
You get A list of what broke, with every item fixed or explicitly deferred in writing.
- 06
Live, and handed over
One training session and a written runbook: what each system does, what it will not do, where it hands to a person, and what to do on the day it is wrong. Monitoring and alerting go on, pointed at us.
From you An hour each for the people who will use it, and a decision on who gets called first.
You get It is live, the build is closed, and running it becomes our job.
The build ends. Then the care plan runs.
The build is fixed scope at a fixed price. When it closes you have the systems running on live work, a written runbook, training done, alerting pointed at a named person on our side, and a list of what we deliberately did not automate and why.
What you do not have is something that keeps working on its own. That is not a flaw in what we built; it is what automation is. Integrations break, models change, and your business does not hold still. So the monthly figure is not an upsell you can decline — it is quoted with the build, so the whole cost of year one is on the table before you commit to any of it.
If we stopped, it would stop. Not that afternoon, but within weeks.
Monthly required, and you pick one of two
Care · Maintain — what the month covers
- Being the one who notices. An API version gets retired, a field gets renamed in your CRM, a model changes how it reads a message. Someone has to catch that before your customer does, and it is not going to be you.
- Fixes, with no ticket price. Breakage is not a change request. Repairing what we built is the thing you are already paying for, not a line item that appears afterwards.
- Small changes as the business moves. A new job type, a new form, a new branch, a new person who needs different access. Businesses move; a system that cannot move with them stops being used within a quarter.
- The tooling, held and operated. AI model usage and any tooling the build needs run on accounts we hold and operate, and are billed through to you at cost — quoted before we start, never marked up. Your own systems and data stay yours.
Care · Continuous — everything above, and
- The next item on the map. Once what we built is stable, we go back to the ranked map and take the next thing on it. The map does not stop being useful the day the build closes.
- The map stays current. Re-measured as the business changes, rather than going stale the day the build closed.
- Agents and skills that grow. The business brain keeps learning your exceptions instead of being finished.
The difference is build capacity, not service level. The floor does not move; Continuous is added to it, not traded against it.
A 60-minute diagnostic, then a fixed scope. Nothing owed either way.
Free, no pitch deck. We work out whether there is anything here worth building — including when the answer is that there is not.