How to pick an AI company in the Netherlands in 2026

Choosing an AI company is not a tool choice

Anyone looking for an AI company right now walks into a market that has multiplied fivefold in two years. Hundreds of Dutch agencies, from one-person operations to consultancy arms of Big Four firms, claim "AI implementation for SMBs". The gap between those who deliver a PowerPoint and those who put a working system in production is a factor of ten in value. This is my honest checklist, written as a builder and calibrated against peer agencies I have run into in the same selection rounds. One thing upfront: the agencies that pitch hardest are usually the worst fit. There is no "book a meeting" button at the top of my site. That is not an oversight, that is a filter.

The seven selection criteria

1. Can they show you what is running in production?

The question. An AI company that only shows pilots, demos, or POCs often does not deliver anything that is still running a year later. Ask concretely: which client has an AI system live now, how many hours per week does it save, and may I speak with that client for ten minutes? An honest agency has three to five examples ready. A marketing agency that does AI as a side product does not. If they hesitate or redirect to "we can demo it", they have nothing in production.

2. Do they build or only advise?

Two camps: builders and strategists. Strategists deliver reports, builders deliver working systems. Both have a place, but SMBs usually benefit more from a builder that can also think strategically than from a strategist who hands you off to a vendor afterwards. Ask who writes the actual code. If the answer is "a subcontractor in Poland", you know where you stand. AI strategy without pilots is not strategy, it is a report.

3. Who sits on the other side of the table?

At larger agencies you get account managers and project managers between you and the technology. At small agencies you talk directly to the builder or the founder. For SMBs direct contact with the founder is faster and cheaper, because intermediate layers cost money and leak information. Ask: who is my fixed point of contact and how often will I speak to that person? If the answer is "a delivery lead" and the engineering sits two layers further out, you are stapling your brief to a relay.

4. Tool-agnostic or partner-locked?

A certified UiPath partner will recommend UiPath. A Microsoft Gold Partner will recommend Copilot Studio. That advice is not necessarily wrong, but it is biased. Ask how they arrive at a tool choice and what their advice would be if you did not want Microsoft or UiPath. An honest agency evaluates per project and has no partner bonus to defend. For the RPA decision see /en/rpa, for agentic AI see /en/ai-agents.

5. What do they do with your data?

Three things to check. Where is your data processed, EU-only or US cloud? Is it used for model training, which is the default on consumer tiers? And what does the DPA say about retention and sub-processors? For healthcare, legal, and financial this is not a detail, it is a dealbreaker. A serious agency shows how they can build on-premise or EU-only when that is required. For the broader AI Act context see /en/ai-act.

6. AI literacy (Article 4) in order?

From 2 February 2025 the EU AI Act requires staff using AI to have sufficient knowledge. An agency that does not explain this, does not deliver training, and does not produce a compliance file per use case exposes you to fines and reputation damage on audit. The signal goes wider: does the agency itself use AI well? If their own people only open ChatGPT to draft a birthday email, you are not going to have an AI system built by people who barely hold the instrument. AI-first at your vendor is a hard filter. See /en/ai-training for our take.

7. What does it cost and how do they bill?

Three models you encounter: fixed price per project, hourly rate, and retainer. For defined pilots fixed price works best. For open-ended work hourly billing pays off, provided there is a budget cap. For ongoing monitoring plus small changes a retainer is logical. The real conversation is not about the rate: how is overage billed, what is the exit clause, and what does the client own (code, prompts, documentation)? Whoever answers vaguely is parking the problem at your end of the table.

Red flags

Three patterns I see in the market and read as warnings.

"Six months of AI strategy, then we start building." For 90 percent of SMB cases that is overkill. A good AI roadmap takes two to four weeks, after that you build. Agencies selling six months of strategy often sell strategy to you and do not build themselves. You become their research budget.

No fixed prices, no scope document. If an agency starts with "we can only give you a price after an extensive intake" and that intake itself costs €5,000 to €15,000, the scope is unclear and you are paid prospect-research. For defined use cases a serious builder gives a target price after a 30-minute conversation. Not exact, but honest.

Tools pitch instead of problem analysis. "We use UiPath and GPT-4 and n8n and LangChain" is a tool stack, not an approach. A good conversation starts with "what does this cost you now, in hours or euros, and what would halving it be worth?". Tools come after. Whoever leads with their stack is telling you what they want to sell, not what you need.

Realistic price ranges

How DataDream prices: no fixed rates on the site. Per project a fixed price after a 30-minute discovery call. Scope varies too much to put a meaningful number on it upfront, and a price without context poisons the conversation. What you can expect: scope first, price second. With every agency you compare, ask for a fixed price on defined work, an overage clause for open-ended work, and who owns code, prompts, and documentation after delivery. Whoever cannot give a clear answer to that in a first conversation will not give a clear answer in a quote either. Schedule a conversation via /en/contact.

How DataDream positions against this checklist

I did not invent this checklist; clients asked exactly these questions during selection conversations. My answers, in the same order.

Production: five+ clients with live AI systems, voice agents, document extraction, content pipelines. References on request. Builder or strategist: primarily builder, strategy only where it sharpens the build. Direct contact: you work directly with me (founder and lead engineer), no account-manager layer in between. Tool-agnostic: no partner lock-ins, per project I evaluate UiPath, Power Automate, n8n, Workato, custom Python and the right LLM provider. Data: EU-only deployment available, on-premise possible for sensitive sectors, GDPR compliant by default. AI literacy: Article 4-compliant training via /en/ai-training, compliance file per use case. Rates: fixed pilot prices for defined projects, hourly for open-ended. Ownership stays with the client, always.

For the full service overview see /en/ai-solutions. For agentic AI and voice AI see /en/ai-agents. For RPA and workflow automation see /en/rpa.

Conclusion: three questions to ask upfront

In every conversation with a potential AI company, these three questions filter out the time-wasters.

One: which client currently has an AI system in production that you built, and may I have a ten-minute call with them? Hesitation or redirection to "we can demo it" is a red flag. Two: suppose I want a working first version in six weeks; which use case would you recommend, and with which tools? A good agency gives a concrete answer within five minutes. A poor agency says "that depends on the intake process". Three: what does your standard data processing agreement say about AI data? If they have one and can explain it, fine. If not, look elsewhere.

Want to apply this checklist to DataDream? Schedule a free 30-minute discovery call. No sales pitch, just an honest assessment of whether we are the right choice for your situation. Or start with the free AI Readiness Scan for an initial direction.