Line one
Technology has to move the result.
The studio’s flagship line, which starts by measuring where you stand, passes through a strategy somebody can execute, and ends in systems that run and are kept running by the same people who proposed them.
The clock
The duty stopped waiting for you.
The AI Act asks anyone who puts an AI system on the market, or uses one, to keep a sufficient level of AI literacy among the people around it. That obligation has applied since 2 February 2025. The training that answers it is the academy, run by role. From August 2025 the Member States were due to name supervisory authorities and set the penalties, and on 2 August 2026 most of what remained of the Act came into application.
So nobody here is being prepared for something that has not arrived. This is work on something already running, and it is done at governance level: which systems are already inside the house, what risk each of them carries, who needs to know what, and where the trail sits when somebody asks for proof.
The services
Twelve doors, and the first one never changes.
| The service | What you receive | The shape |
|---|---|---|
| Digital maturity diagnosticTHE ENTRANCE | A scored profile, a map of the opportunities, and the first sketch of a roadmap ordered by value. | Fixed price, two to three weeks |
| Digital and AI strategy with a roadmap | A roadmap of twelve to twenty-four months, costed initiative by initiative and sequenced the way the dependencies allow. | Fixed scope |
| Where AI is worth the money, ranked | Your processes read against what the technology can honestly do this year. Every case is scored and gets a verdict: build, buy or wait. | Fixed scope |
| AI agents: design, build, operate | Agents that do real work in service, documents, research and on the shop floor, answering out of the technical file or the quality records, with measurable goals, evaluations and a human in the loop. | Proof of concept, then build, then a monthly retainer |
| Search across your own documents | Answers drawn from your documents and your systems, with evaluations and guardrails. The right approach for most enterprise cases. | Fixed scope or sprints |
| Process digitisation and automation | The process is redesigned first and automated second, where it pays: an order from the quote to the loading bay, production planning, the join between the ERP and the warehouse, approvals and reporting. | Sprints |
| Data foundations and business intelligence | A data estate that has been put in order, and dashboards that answer the questions a board asks rather than the ones a tool offers. | Build, then a monthly retainer |
| Customer channels and digital products | Sites, shops, portals and products, designed and built by the studio at the studio’s own standard. | Fixed scope or sprints |
| Getting it adopted, and training the people who will run it | Role-based training, a network of people inside who carry it on, and measurement of whether the new thing is genuinely used. | Fixed scope |
| AI governance and readiness for the Act | Classification of the systems you already run, transparency duties, a literacy programme, written policy, and a trail you can show. | Fixed scope |
| Private AI infrastructure, managed | Open-weight models deployed and kept running on our infrastructure or on yours, for the cases where the data has to stay where it is. | Setup fee and a monthly retainer |
| Fractional Chief Digital Officer | The person who leads the transformation, on a monthly retainer, and reachable between board meetings. | Monthly retainer |
Where the model lives
Three places, all three first class.
- A frontier provider, over its API. The fastest road to something working, with advice on where the data should sit inside Europe.
- An open-weight model on the studio’s infrastructure. We run it, we watch it, and the bill stops growing with every request.
- The same model on yours. For when the data must not leave the house, because a regulator, a contract or plain sense says so.
Faith has nothing to do with that choice. The studio does not make it alone either. We work it out with you against the volume that will really pass through the model, and we write the arithmetic down so it can be checked later.
The order
How a line runs.
One method sits under every engagement and it is not repeated here: it is written out on the consulting page. What follows is the order inside this line, from the first week to the monthly retainer.
The questions
What people ask before the first meeting.
- We have no idea where to start. Then you start with the diagnostic, which exists for exactly this conversation. After it the decision is no longer taken blind.
- What does it cost? Each service above names its own shape in the table beside it, and what moves a price is written out on the pricing page. Say what the task is and the band comes back in writing.
- Do we have to buy our own machines? Usually not. We work out where the line falls for your volume, and if it is far away we stay with a provider.
- What if it does not work? That is why cases are scored before they are built rather than after. If the portfolio shows nothing pays for itself, you will hear that from us, and it counts as a result.
- How much of my people’s time does this cost? For the diagnostic, eight to twelve conversations of about an hour, with the leadership, with IT and with the people who do the work itself, plus read access to the systems in question. Nothing on the floor stops for us.
- Who does the work? The people you spoke to. The studio does not sell hours into somebody else’s team and does not send a substitute for the person who made the promise.
The beginning
Start with the diagnostic.
Two things are worth putting in the letter: the process that eats the most time, and what you have already tried on it. A person answers.