Service · Web development

Developing an AI strategy for your organisation.

You know AI is going to change something in your market, but not yet what, in what order, or with what risks. We build an AI strategy that takes your data, your people, your regulations and your budget into account. And we don't walk away after the first presentation: we help build it too.

Readiness assessmentOpportunity mappingEU AI ActBuild vs buy

An AI strategy is not a list of tools, it is a choice of which problems to tackle first.

Most organisations that come to us have the same story. Management wants "something with AI" to happen. One department is already experimenting with ChatGPT, another team has quietly started a Copilot pilot, and the CIO is worried about data leaks. Everyone has an opinion, nobody has the overview. The question you actually want answered is: where will AI deliver measurable value for us in the coming period, without steering us into legal or reputational risks?

A good AI strategy gives a well-founded answer to that. Not as a PowerPoint full of buzzwords, but as a concrete roadmap that takes your people, your infrastructure and your regulatory context as its starting point. We create that strategy and, what sets us apart, we then help build the first pilots. Our advice isn't handed over to an implementation partner; we get our own hands into the system. That keeps us honest too: an adviser who never writes code themselves can far more easily recommend something that doesn't work.

An AI strategy differs from a digital transformation programme. Digital transformation is broadly about technology, processes and people, whereas an AI strategy zooms in on where machine learning, language models and agents can add value, and which prerequisites are missing in your organisation. Many clients run both programmes in parallel or in sequence, especially if their broader digital maturity is still limited.

Three tracks we work through together.

Depending on where you stand, an AI strategy starts with sharper insight, a focused roadmap, or building the first pilot straight away. We often work through them in that order, but not always.

Track 1 · fixed sprint budget

AI readiness assessment

We map out where your organisation stands today: data maturity, infrastructure, team skills, governance and culture. Not a questionnaire with scores from 1 to 5, but conversations with IT, business owners and management, combined with a short analysis of your actual data and systems. The result is an honest picture of what is feasible and what needs fixing first. Data quality, accessibility or organisational ownership often turn out to matter more than the choice of model.

Data auditSkills mappingInfrastructure scanGovernance baseline
Track 2 · fixed sprint budget

Opportunity mapping and roadmap

Which processes or products could benefit from AI, and which could not? We score opportunities on three axes: potential value (cost savings, revenue growth, customer experience), implementation complexity and risk. From that overview follows a phased roadmap: which pilots we start with for quick wins, which larger transformations we save for later, and which ideas we drop because they promise more risk than return. We also attach an initial order-of-magnitude investment estimate so that management can translate the roadmap into budget discussions.

Use-case workshopsValue scoringRisk mappingRoadmap
Track 3 · fixed sprint budget

Pilot implementation

We help build the first concrete application from the roadmap. That could be an AI agent that takes over a specific process, an internal knowledge assistant for your staff, or an extension of your existing software with AI functionality. It is not meant as a finished product, but as proof that the strategy works in your reality. During this phase we sharpen the strategy based on what we learn, because theory and practice diverge more often in AI projects than in traditional software implementations.

Proof of conceptLive pilotMeasurement frameworkKnowledge transfer

What you get at the end.

An AI strategy that sits on the boardroom table, with enough detail underneath to take the first steps the next morning. No 80-page report that nobody opens.

  • Readiness reportAn honest assessment of your data, infrastructure, team and governance, including what needs fixing before AI can land in a meaningful way.
  • Use-case portfolio with scoringTen to twenty concrete AI applications scored on value, complexity and risk, so the choice of which to tackle first is not made on gut feeling.
  • Phased roadmapA concrete plan spanning several quarters, with dependencies, milestones and go/no-go moments. Not a Gantt fantasy, but a workable plan.
  • Risk register and AI Act categorisationFor each use case, the EU AI Act risk classification (minimal / limited / high / unacceptable), plus GDPR implications and ethical pitfalls.
  • Build-vs-buy adviceFor each use case: build it yourself, buy (Copilot, ChatGPT Enterprise, Claude Enterprise, Gemini), or a hybrid. With supporting reasoning, not just "we recommend X".
  • Governance blueprintWho decides on new AI projects, which guardrails apply to data and models, how we measure success and when we pull the plug.
  • Pilot implementation (optional)We help build the first use case from the roadmap, so the advice is tested in your reality straight away and does not stay stuck in a PowerPoint.

When an AI strategy is the right choice.

Four signals we often see in organisations that are ready for a strategy engagement. If you recognise one or more, we would be glad to talk further.

Management

"Something with AI" is on the agenda

The board expects action, but there is no plan. You want to avoid AI turning into a collection of isolated experiments with no shared direction. The C-suite needs a sound basis for budget allocation.

Shadow AI

Everyone is already running pilots, with no coordination

Different departments are experimenting with ChatGPT, Copilot or industry-specific AI tools. Nobody has visibility of where data ends up, which contracts have been signed, or what is actually being delivered.

Compliance

The EU AI Act is coming your way

You want to know in advance which planned AI applications fall into high-risk categories, what documentation you need to keep, and how to organise that in practice, before the regulation catches up with you.

Competition

Competitors are making claims you need to be able to put into context

Market players claim impressive AI results in pitches and webinars. You want to be able to tell marketing from reality, and decide which claims you can make yourself without damaging your credibility.

Not sure about a large project yet?

Test your idea first: a working prototype in 1 day

With OneDayBuild, we make your idea tangible in a single day for €1,150, so you know whether further development is worth the investment. Decide to go ahead with the full build? Then we deduct the cost in full.

View OneDayBuild →

How a strategy engagement works.

1

Introduction

A conversation in which we understand where your question comes from, who is at the table, and what decision you want to be able to make based on the strategy. No sales pitch: we will also tell you honestly if you are not yet ready for a strategy engagement.

2

Readiness and opportunity mapping

Interviews with IT, business owners and management. A short data audit on a sample of your systems. Workshops to surface use cases. At the end of this phase: a shared view of what is feasible and which opportunities look most promising.

3

Roadmap and governance

We build the roadmap with dependencies, risk classification and build-vs-buy choices for each use case. At the same time, we work out the governance structure: who decides, which guardrails apply, and which yardstick is used. One closing session with management to secure commitment.

4

Pilot and ongoing support

We build the first pilot from the roadmap, or support your own team in doing so, whichever suits you. After that, we remain available for the next hurdle: new use cases, AI Act documentation, or a revision of the strategy when the landscape shifts again.

Why organisations work with us.

Four things we do differently from the average strategy consultancy, which we spell out so you know what you are buying.

Advice + build

We help build, too

Our background is software engineering, not PowerPoint. A strategy we deliver is executable because we know what execution involves, and because we often build the first pilot ourselves. That keeps the advice honest and grounded.

Vendor-neutral

No preference between Claude, GPT, Gemini or open source

We have no partner contract pushing us towards a single AI vendor. For each use case, we choose on merit: reasoning quality, data residency, cost model, or a multi-vendor setup to avoid vendor lock-in.

Realistic

We also tell you where AI adds nothing

Not every process is improved by AI. Sometimes a good search function is more effective than a chatbot. Sometimes workflow automation without a language model is faster, cheaper and easier to explain. We state this explicitly in the roadmap.

Dutch context

Focused on Dutch regulation and language

The EU AI Act in practice, GDPR implications of American AI vendors, sector regulators (DNB, NZa, AP), Dutch-language model performance. This is built into our engagements as standard, not as an appendix but as a design parameter.

Frequently asked questions.

What decision-makers ask us before an AI strategy engagement begins.

When does an AI strategy make sense for our organisation?
An AI strategy makes sense as soon as you move beyond isolated experiments. Three signals: management wants to make well-founded decisions about the AI budget, several departments are already running pilots without coordination, or you fall under regulation (EU AI Act, sector regulators) that requires documentation of AI systems. If you are still entirely in the exploratory stage, you are better off starting with a short readiness scan or a focused workshop than with a full strategy engagement.
How long does an AI strategy engagement take?
A readiness scan often runs over a few sprints. A full strategy, including opportunity mapping, roadmap and governance blueprint, takes an engagement of several sprints. We work iteratively: you don't have to wait until the end to see something, because every phase delivers tangible output you can build on in the meantime. The timeline depends mainly on how quickly your people are available for interviews and workshops.
Are you consultants or builders?
Both. Our background is in software development, and we have been building AI agents, web applications and integrations for years. That means we write strategies that are executable because we know what execution involves. The difference from classic consultancy: after the strategy, we can build the first pilot ourselves or work closely with your own team. For heavy enterprise implementations, see our page on enterprise AI implementation; for a specific agent, see building an AI agent.
How do you handle the EU AI Act and GDPR?
In every strategy, we classify use cases according to the EU AI Act's risk categories (minimal, limited, high, unacceptable). For high-risk systems, we map out the documentation obligations: technical file, conformity assessment, post-market monitoring. We assess GDPR implications per use case, especially around automated decision-making, profiling and data transfers to AI suppliers outside the EU. For deeper compliance questions, we work with lawyers or refer you to our page on AI Act compliance software.
Which AI providers do you recommend: Claude, GPT, Gemini or open source?
We are vendor-neutral. We have built working solutions on Claude, GPT models, Gemini and open-source models such as Llama and Mistral. The choice depends on the use case: how critical reasoning quality is, what the data residency requirements are, how much volume will run through it, and whether a vendor-independent architecture matters. In the strategy, we make these choices explicit per use case, rather than locking in a single provider up front.
What if a pilot from the roadmap fails?
Some pilots will fail. That is part of the process, and we design for it. Every pilot gets clear success criteria and a go/no-go moment up front. If a pilot fails, we document why (technology, data, adoption, business case) and decide whether it is worth trying again or dropping the use case. Failure is only a problem if you learn nothing from it; in the governance structure, we define exactly how that learning happens. Pilots with no learning value are more expensive than failed pilots.
What about the costs?
The price of a strategy engagement depends on its breadth (one department or organisation-wide), the depth of the readiness scan, and whether you also want pilot implementation included. We have no fixed price list and no indicative tier prices, as these would mislead you more than inform you. In an initial conversation, we outline which option fits and roughly what that means in terms of investment. We are no more expensive than the large consultancy firms, and unlike them, we also build with you afterwards.
Do you also train our team?
Knowledge transfer is part of our engagements as standard: your people should be able to carry on with the strategy without being permanently dependent on us. For broader AI training for your employees (prompt engineering, responsible use, sector-specific applications), see our page on AI training for companies. A strategy and well-trained people are two sides of the same coin.

Talk to us about your AI strategy.

A half-hour introductory conversation, with no obligation. We listen to where your organisation stands and where the pressure is coming from, and give you direction you can use straight away, even if we turn out not to be the right partner for what comes next. An AI strategy is often tied to broader choices, so we are happy to discuss how it relates to your digital transformation programme.

Edit content