The seat: the top tier of Run
Who owns this after we leave?
For most companies the answer is their own team, which is what everything we build is designed for. For some it is not enough: they do not have one job worth building, they have a queue, and nobody whose job is to put it in order and say what ships next. A fractional AI officer is that person, inside your team, month to month.
How it runs
Monthly
Blueprint
EUR 4,900Prioritised use cases and a roadmap, on your real processes.
You decide
You pick what ships first. Nothing is open-ended.
Build
The queue moves every month, inside your systems.
Handover
Your team takes over whatever is live.
Your company
Ships AI without hiring for it.
Small start, then a queue that actually moves.
This is the largest version of Run, not a separate product.
Every build we do ends the same way: your team is trained, you own the system, and it keeps working whether or not you ever speak to us again. Run is the optional step for companies that would rather not carry it themselves, and it starts far below this. Monitoring and exception handling on one system is the small end. The seat is the large end, where somebody senior is also deciding what gets built next.
You do not have to buy a seat to have a system watched, and you should not buy one before there is a queue worth running. The diagnosis is the honest way to find out which of those you are, and its fee is credited against the first month if you go ahead within 90 days.
What the seat does
- Owns what gets built next, and keeps that order current as the business changes
- Runs the build queue inside the systems you already have
- Watches what is already live, and handles the exceptions before they become incidents
- Trains your team on everything that ships
- Reports against the measures you agreed, not against activity
- Available to your team for the questions that used to stall things for a week
Month one
A paid pilot, against one named thing.
Before month one starts we write down the one deliverable it will be judged on. If it lands, the seat continues month to month. If it does not, there is no second month, and you keep whatever was built along the way.
A seat is the easiest thing in this business to sell and the hardest to prove, so it seemed right to make the first month prove itself.
Who this fits
A queue, not a job
Several things worth building, and an order that matters more than the individual pieces.
Owner-led and growing
Roughly 20 to 250 people, where the founder or MD is still the bottleneck on anything new.
No AI lead internally
Nobody whose job is to decide what to build, in what order, and to say honestly whether it worked.
If you already know the one thing you want built, buy that instead and run it yourself afterwards. The five jobs are on what we build, and we will say on the call which of the two you actually have.
Questions buyers ask
Do we have to commit?
No, and the first month is structured so you do not have to guess. Month one is a paid pilot against one named deliverable that we agree before it starts. If that deliverable is missed, there is no second month and the engagement ends cleanly. After that it is month to month, billed in advance.
How is this different from hiring someone?
A senior AI lead is expensive, slow to find, and hard to assess if AI is not already your field. This is the same seat immediately, at a fraction of a salary, with a team behind the person. You are buying judgement plus delivery capacity rather than headcount.
We only have one job we want built.
Then you do not need this. Buy the build, take the handover, and run it yourself. Your team is trained for exactly that and the system is yours. This page exists for companies with a queue rather than a job, and we will say which one you have on the call.
Is this the only way to get support after handover?
No. This is the largest version of the Run step. Smaller ones exist, from monitoring and exception handling upwards, and they are priced against what you actually need in the conversation. Nobody has to buy a seat to get a system watched.
Who actually does the work?
The people you meet. Mapular AI is a small Berlin team that ships production systems, and the person in your sessions is one of the people building. There is no account layer and no handover to a delivery team you have not met.
What do we own?
Everything: the code, the data, the accounts, the documentation, and the plan itself. Your team is trained on each thing that goes live. If the engagement ends, nothing stops working and nothing has to be handed back.
What does it cost to stop?
Nothing beyond the month you are in. Billed monthly in advance, continuing month to month, so the queue has to keep earning its place.
Find out in 25 minutes whether you need a seat at all.
Bring the list of things you keep meaning to get to. If it is one thing rather than a queue, we will tell you, and that is a cheaper conversation for you.