Services

Measure what your AI actually does, then fix what the measurement reveals.

We do one thing, and it is a discipline rather than a task: making measurable what almost everyone validates on vibes. Hence the order we work in. Evaluation tells you whether your system holds and where it breaks; it almost always points to the same place, your data and the rules nobody wrote down; we then turn those into assets that are structured, documented and defensible. Pragmatic and vendor-neutral: we work in your stack, with your models and your cloud.

Book an AI-readiness assessment

How engagements start

Most partners want to sell you a six-month build before they understand your business. We don't. Nearly everyone starts with the AI-readiness assessment: our first engagement, fixed scope and fixed price (€10,000 excl. VAT, typically one week). It looks at the three things that decide everything else: your data (what it's actually worth and what's missing from it), your architecture (what your stack, your security constraints and your cost envelope genuinely allow), and your business (which use cases pay off, which can wait, and in what order to take them on). You leave with your AI Roadmap. From there, getting your data and your expertise AI-ready run in whichever order your roadmap prioritizes, after evaluating your system if one already exists; coaching layers on top of either. Only then is the rest of the work scoped to exactly what it takes. The 30-minute call before it is free and carries no obligation.

Download the free guide to evaluation (PDF)

AI Readiness Assessment

The front door · low commitment, high clarity

What it's for

You know AI should matter to your business, but you don't know how far your data is from being usable, what your existing architecture can carry, or which use cases are genuinely worth it. This is where to start.

What we do

Your data: where it lives, what it's worth, and how far it is from AI-ready.
Your architecture: what your stack, your security constraints and your cost envelope actually allow, and what would have to change.
Your business: which use cases pay off, which can wait, and in what order to take them on.
Surface the risks (technical, data, and organizational) before you spend on them.

What you leave with

Your AI Roadmap: the gap to AI-ready, the target architecture, the prioritized use cases, and what each step would take. Yours to keep, whether or not you continue with us.

Fixed scope · €10,000 excl. VAT · one week, typically.

Book an assessment

Your AI system → evaluated

The first piece of work · short, and it reveals the rest

What it's for

Something is already running and nobody can say whether it answers correctly. The pilot doesn't reach production because no number defends it, and every opinion in the room carries the same weight. This is where to start once the system exists.

What we do

Build a set of cases drawn from your reality: your actual records, real user questions, failures included.
Write the criteria case by case: what a correct answer is at your company, and what proves it.
Calibrate an automated judge against your labels, and measure its agreement before trusting it at scale.
Industrialize the harness so it replays on every change of prompt, index, data source or model.

What you leave with

An evaluation harness you own: cases, criteria, a calibrated judge and replayable results. It becomes your regression test, and it names the data to fix first.

Scoped from your AI Roadmap · on a system already in service.

Book an assessment

Proprietary Data → AI-Ready

The core work · the part most consultancies skip

What it's for

Your data is your moat, but it's raw. Locked in silos, inconsistent formats, documents, and tacit expertise a model can't reason over. Before AI can deliver, the foundation has to be right.

What we do

Structure and model your data and domain knowledge into well-defined, documented assets.
Turn scattered records, documents, and processes into something a model can use reliably.
Deliver data kept current, with quality and lineage you can trust.

What you leave with

Proprietary data that's genuinely ready for AI (documented, structured, and defensible), plus the foundation every future AI use case will stand on.

Scoped from your AI Roadmap · 1 to 3 weeks, typically.

Talk about your data

Expertise → AI-Ready

The knowledge that isn't written down anywhere

What it's for

Your moat isn't only in your tables. It's how your people decide, the rules they carry in their heads, the processes no document fully captures. A model can't use what was never written down.

What we do

Elicit and formalize the domain expertise your team takes for granted.
Turn tacit rules, judgments, and processes into explicit, structured knowledge.
Encode it as constraints, logic, and knowledge graphs a model can reason over and you can verify.

What you leave with

Your domain expertise captured as a documented, verifiable knowledge asset: no longer trapped in people's heads, and ready for any AI to build on.

Scoped from your AI Roadmap · 2 to 6 weeks, typically, depending on stakeholder availability.

Talk about your expertise

Your teams → autonomous with AI

Coaching · so it lives on after we leave

What it's for

AI-ready assets only pay off if your people can use them. You want decision-makers who know where AI is worth it, and technical teams able to run and extend the foundation, without depending on us.

What we do

Coach your decision-makers: picking use cases, framing risks, driving adoption at your pace.
Train your technical team to operate, maintain, and extend the data assets we build.
Install the practices (quality, documentation, governance) that keep the foundation healthy over time.

What you leave with

An autonomous team: decision-makers who know where to invest, engineers who run the foundation without us.

Scoped from your AI Roadmap · standalone or alongside the data and expertise engagements.

Talk about coaching

Is this for you?

We're not the right choice for everyone.

This is for you if…

You own proprietary data and know-how built up over years: your edge is what you know.
You want AI that runs in your stack, with your models and your cloud, with no vendor lock-in.
An AI pilot isn't reaching production, or you want to make sure the first one doesn't stall.

This isn't for you if…

You're shopping for a turnkey chatbot licence: we build assets, we don't resell tools.
You want developers on a day rate: we sell an outcome, not staff augmentation.
You don't yet have data or expertise of your own: there's nothing to make AI-ready.

Frequently asked

What we get asked before starting.

How long does the assessment take?

One week, typically, at a fixed scope. You leave with your AI Roadmap no matter what.

How much does it cost?

The assessment is €10,000 excl. VAT, at a fixed scope: no day rates, no overruns. The work that follows (evaluation, data, expertise, coaching) is then scoped from the roadmap, never sold blind.

Does our data have to leave our environment?

No. We work in your infrastructure, with your models and your cloud. Your data doesn't have to leave: that's a principle, not an option.

Do we need an in-house technical team?

No. Many of our clients have no IT department. We do the work, we document it, and the coaching engagement can make your team autonomous on what we build.

What happens after the assessment?

The AI Roadmap is yours to keep, whether or not you continue with us. If you do, each engagement is scoped from its priorities: evaluation, data, expertise, coaching.

Not sure where you are?

Start with the assessment.

It's the fastest way to know how far your data is from AI-ready and what's worth cleaning first. Fixed scope, fixed price, and your AI Roadmap is yours to keep either way.

Book an AI-readiness assessment