Agents that read your documents, reconcile your data, prepare the next action, and stop where a person must decide. We design them, plug them into your information system and run them with your teams, until they no longer need us.
The business leads, the system follows.
AI where it is useful and nowhere else.
Teams that run without us once we leave.
Depending on your situation, we advise, we build, or we acquire the company. Same method, same standards, same limits: a person decides, and AI comes in only where it is useful.
Assessment of the existing system, mapping of flows and friction points, a rebuild roadmap ranked by business value.
We settle what has to be rebuilt, what has to be replaced, and what should simply go.
We design and ship: data integration, process automation, supervised AI agents, business interfaces.
We stay until the in-house team runs it alone, on a schedule set at scoping.
We acquire sound companies whose information system caps growth or margin, and we run them ourselves.
The rebuild then becomes a shareholder’s commitment, not a service contract.
We do not bill time without a defined scope. Each format has a duration, a price set in advance and a deliverable you can use without us.
A few weeks, fixed fee. A shared assessment with management, a ranked and costed roadmap, a first project scoped.
A scope, a schedule, a price, written at scoping. Delivered in stages, documented, with a planned handover. The next one is decided in light of the previous one.
If you are preparing a succession, a separate framework, with your usual advisers. We only raise the subject if you do.
Four steps, in this order. Each one is useful on its own, even if you stop there.
We watch how the work actually gets done: the side spreadsheets, the re-typing, the manual checks. That is where the cost of the current system hides.
AI amplifies whatever it is given. We start with reference data, business rules and reconciliations, so that automation rests on solid ground.
Every automated action is designed to be traced, reversible and documented. Teams see what the system proposes and keep the final say, especially where customers or care recipients are involved.
We set the indicators at scoping and we train the team to keep the system alive without us.
On doubtful data and blurry processes, it produces mistakes faster. On solid ground, it changes the scale of what a small team can do.
A reconciliation, a consistency check, an export: most of the friction in a system is handled by simple, readable, verifiable rules. We keep models for the cases where volume and variability go beyond what a rule can do.
Every automation that directly concerns a person, customer or care recipient, goes through human validation. That is not only a regulatory requirement. It is a condition of trust.
Acquiring companies forces us to live with our own advice. For us, that is the proof that we recommend nothing we would not be ready to run ourselves.
One billed hour, several payers, a matching puzzle rebuilt every month in a spreadsheet. We describe the system we propose, with a worked example: what AI agents prepare, what a person decides, what the system will never do.
No. It is the usual situation of the companies we work with, and our services are designed for it. We work with management, finance and the people who run operations. If an IT provider is already in place, we coordinate with them.
Rarely as a first step. Much of the value comes from what happens around the software: the data it is given, the checks before entry, the reconciliations and the exports. We only recommend a replacement when it is justified.
It depends on the context and on how sensitive the data is. We favour environments hosted in the European Union and providers that do not use your data to train their models. The choice is documented at scoping and validated with you.
An assessment takes a few weeks. Delivery projects are sliced so that a first result lands quickly, then proceed step by step based on what we observe. We prefer several short steps to one long project.
The assessment is a fixed fee, set before we start. Delivery projects are priced one by one at scoping, with a written scope and schedule. We do not bill time without a defined scope.
Clearly, and from the start. A consulting or delivery engagement is not a prelude to an acquisition: we only raise the subject if you do, and it then gets a separate framework, with your usual advisers (lawyer, accountant). Information gathered during an engagement serves that engagement only.
We acquire, on our own account, alone or with industrial partners, small and mid-sized companies whose business is sound but whose information system limits growth or margin. We then run them and lead the rebuild from the inside, over time.
Thirty minutes are enough to know whether we can be useful. If we cannot be, we will say so.
Two lines will do: your organisation and what is on your mind.
Long enough to understand your situation, your constraints and what you expect from us.
A few weeks, a scope and a price fixed in advance, and a first project scoped at the end.