About
An AI consultancy that stays until the thing is used.
We advise on AI strategy, build prototypes that prove or disprove an idea, and do the adoption work that decides whether any of it survives contact with the organisation. Those three things are usually split across suppliers who never speak to each other.
Background
Where we come from
Our work spans applied machine learning and generative AI delivery, the strategy and change work around it, and European research programmes. We have built and shipped production AI systems, run the discovery and roadmapping that decides what gets built, and led work packages inside funded R&D projects.
That combination shapes how we advise. We know what a model can realistically do, what it costs to run once real traffic hits it, and what happens six months later when the people expected to use it have not changed how they work. Those perspectives are rarely in the room at the same time.
We work across Europe, with roots in Brussels and Athens. We do not name clients or projects publicly - but we will talk through comparable work in detail, under NDA where that is needed.
How we work
Four things you can hold us to
01
Every recommendation carries a number
What it costs to run at your volumes, what it returns, when it pays back, and which assumption the case depends on. An AI proposal without economics is a hypothesis, and we do not present hypotheses as advice.
02
Adoption is in scope, not handed off
A capability nobody uses returns nothing. Operating model, workflow, training and adoption measurement are part of the work rather than someone else's problem after go-live.
03
We say when the answer is no
A use case that will not pay for itself, a prototype that disproved the idea, a call not worth entering - these are useful results, and we deliver them as readily as the other kind.
04
Fixed scope wherever possible
Most of our work has a defined deliverable and an end date. Retainers exist for the work that genuinely recurs, not to keep the meter running.
Want to know if we have done this before?
Ask. We will talk you through comparable work in detail, including what did not go to plan - without naming anyone.