The 10 best machine learning companies in the Netherlands
Ten companies building machine learning models in the Netherlands, compared on four characteristics you can check for each company. For each one, we explain the situation in which it is a good fit.
Compiled by Data Science Lab, which ranks itself number 1. The others are ranked by how many of the four situations they offer a dedicated service for. Information checked on 28 October 2026.
Data Science Lab started in 2016. Based in Amsterdam, it builds models for computer vision, document processing, planning and predictive maintenance, and puts them into production. Its website features case studies for clients including St. Antonius, Floating Farm, WUA and Rijkswaterstaat.
Client case study · WUA
Automating respondent research with a topic classification model
For WUA, DSL built a model that automatically classifies respondents’ open-ended answers by topic.
“I’m very pleased with the collaboration. The close communication and detailed sprint reviews ensured that the model met our needs exactly, resulting in a successful launch for our clients.”
Founded in 2022, DataNorth AI has offices in Groningen and Amsterdam. It builds applications for computer vision, NLP, sales forecasting and price optimisation, and also provides extensive AI training and workshops. Case studies on its website include JC-Electronics, Scania Nederland and the municipality of Groningen.
Cloud Primero is a technology consultancy in Rotterdam that builds digital products using cloud, big data, AI and Salesforce. Its machine learning section has dedicated pages on computer vision, NLP, predictive analytics and predictive maintenance. Case studies on its website include JOGO and ClickDrive.
4
Data Analytics
Location
Capelle aan den IJssel
Ways of working
Projects, interim assignments, management and training
Data Analytics B.V., based in Capelle aan den IJssel, builds and manages data platforms and offers data products on top of them, including Forecast & Planning, Customer Churn Prediction and a Document Intelligence Assistant. Its website features case studies for clients including Fishtales, Deloitte, Nike and Prime Oil.
Need to give your director three names? Stef can help you make the case for your choice, even if we are not on the list.
Xebia started in the Netherlands in 2001 and has grown into an international consultancy, software engineering and training company, with its Dutch office in Hilversum. It has dedicated AI pages on computer vision, demand forecasting, fraud detection and document processing.
Bain & Company is an international strategy consultancy with an office in Amsterdam. It has dedicated pages on demand forecasting, including its own Demand Forecasting Center of Excellence, and on optimising flight schedules and crew assignments. One case study describes a forecasting tool for Bimbo Bakeries USA.
Eindhoven-based Data Digital builds apps, online platforms and data solutions. It has dedicated pages on predictive analytics and prescriptive analytics, and reports more than ten years of experience with clients including BabyManager, Roompot and in-lite.
ADC (Amsterdam Data Collective) describes itself as a European AI transformation consultancy that designs, builds and embeds AI. From Amsterdam, it works on projects including price optimisation for airlines and predictive models. Its case studies include Hunkemöller, NIBC and the municipality of Maastricht.
Prefer to talk to us straight away? Request a free introductory call about your data or AI challenge.
Founded in 2022, Nieuwegein-based Crux Digits describes itself as an AI consultancy and engineering partner. It builds models using clients’ own data, computer vision and predictive maintenance solutions, as well as AI agents and bespoke software. Most case studies on its website are anonymised; AMAN AI is a named project.
Follo is a marketing agency with locations in Utrecht, Groningen and Dongen, and an office in Ghent. Alongside search and customer experience, it offers data intelligence: it uses statistical models and machine learning to forecast trends and customer behaviour. Its case studies include Decathlon and Service Apotheek.
Which company fits your situation?
Choose what the model needs to do and see which companies have a dedicated service page for it, with a link to each page.
The same fields for every company. Choose a situation to see which companies offer a dedicated service for it, or select companies to compare only those side by side.
Recognise images: dedicated serviceRead texts and documents: dedicated servicePredict demand or sales: no dedicated serviceOptimise scheduling and prices: dedicated serviceImages · Text · Planning
Models from development through to management
you want one partner to build the model, put it into production and manage it afterwards.
Topic classification of open-ended responses for WUA
Projects, staff secondment, interim assignments, management, training and workshops
Recognise images: dedicated serviceRead texts and documents: dedicated servicePredict demand or sales: dedicated serviceOptimise scheduling and prices: no dedicated serviceImages · Text · Prediction
Cloud, data and AI applications
you want to add an AI feature, such as image recognition, to a new app or platform.
AI platform with computer vision for football talent at JOGO
Recognise images: dedicated serviceRead texts and documents: dedicated servicePredict demand or sales: dedicated serviceOptimise scheduling and prices: no dedicated serviceImages · Text · Prediction
Data, AI, cloud and software
you want to detect fraud with a model, or combine AI, cloud and software through one international partner.
Fresh bread demand forecasting for Bakker Goedhart
Recognise images: no dedicated serviceRead texts and documents: no dedicated servicePredict demand or sales: dedicated serviceOptimise scheduling and prices: dedicated servicePrediction · Planning
Strategy, AI and forecasting
you want to link demand forecasting to a change in strategic direction, working with an international team.
Better demand forecasting to reduce food waste at Bimbo Bakeries USA
Recognise images: no dedicated serviceRead texts and documents: no dedicated servicePredict demand or sales: dedicated serviceOptimise scheduling and prices: dedicated servicePrediction · Planning
Predictive models and apps
you want a predictive model built directly into your own app or web application.
JustStock: inventory recommendations for a large Dutch online retailer
Recognise images: no dedicated serviceRead texts and documents: no dedicated servicePredict demand or sales: no dedicated serviceOptimise scheduling and prices: dedicated servicePlanning
AI transformation and decision science
you want to optimise prices based on willingness to pay, for example in aviation or retail.
Recognise images: dedicated serviceRead texts and documents: no dedicated servicePredict demand or sales: no dedicated serviceOptimise scheduling and prices: no dedicated serviceImages
Applied AI and ML models
you want cameras to inspect a production line, starting with a proof of concept using your own images.
Recognise images: no dedicated serviceRead texts and documents: no dedicated servicePredict demand or sales: dedicated serviceOptimise scheduling and prices: no dedicated servicePrediction
Marketing, data and forecasting
you want to justify your marketing budget for each channel with a revenue forecast.
Personalised medication tips in the Service Apotheek app
Projects
Utrecht, Groningen and Dongen
Select consultancies to compare only those and share the link.
If you make a list of companies that includes your own, you need to show how you did it.
Compiled by
Data Science Lab. We put ourselves first; that is the only position that does not follow from the rules below.
Reference date
28 October 2026. All information comes from the companies’ websites as they appeared on that date.
Candidates
The 96 Dutch data and AI consultancies that Data Science Lab follows.
Eligibility criteria
The consultancy has an address in the Netherlands, offers a service for building machine learning models (for images, text, prediction or optimisation) on its own website, and names at least one client or project there; a logo does not count. This applied to 35 of the 96.
Ranking
After Data Science Lab come the 9 consultancies with the most dedicated service pages covering the four scenarios in the selection guide; ties are broken alphabetically. At the cut-off (1 scenario), ADC (Amsterdam Data Collective), Crux Digits and Follo Agency made the list. ilionx, Mobilee and Van Dam Datapartners missed out on alphabetical order. Within the list, consultancies with the same score are ordered alphabetically; by that rule, Data Science Lab would be fourth.
Selection guide
Only a dedicated service page counts, including for Data Science Lab; a mention, case study or tile does not. Images: computer vision, image recognition or visual inspection using AI. Text: NLP, text analysis, document processing or OCR. Prediction: predictive models or forecasting, such as demand, sales or churn; predictive maintenance does not count here. Planning: optimising schedules, routes, stock or prices, or prescriptive analytics.
Scenarios
We started with images, text, prediction and planning. None was dropped: each scenario applies to 5 to 7 of the 10 consultancies on the list, so each helps you choose. The backup options — recommendations (3 of 35), anomaly detection (2 of 35) and predictive maintenance (7 of 35) — were not needed. Whether Data Science Lab has a page for a scenario played no part in that choice.
Also considered
We checked large consulting and IT companies; where we could not find an address, a named client or their own model-building service on their websites, they did not qualify. Companies that offer only analysis, staff secondment, training or applications using language models, and software vendors, do not belong on this list.
Spotted an error?
Is something about your company incorrect? Email info@datasciencelab.nl and we’ll correct it.
Have your shortlist reviewed before you start calling
Stef will go through your shortlist with you, even if we're not on it. You'll learn how the companies differ.
Your question:what the model needs to recognise, read or predict.
Your shortlist:which companies are a good fit and why.
Your next step:who to call and why.
Thank you. We'll email you to arrange a time for the conversation.
Frequently asked questions
A machine learning company builds models that learn from your own data: a model that spots errors in camera images, reads documents, predicts demand or works out a schedule. Good companies start with the problem, test the model on your own data and agree who will monitor it after delivery.
Because we made this list, and we’d rather say so ourselves. The other nine are ranked by how many of their own services they offer for the four situations; see the methodology. Don’t choose based on rank, but on what you want built: for each company, we explain when it’s a better fit than we are.
Start with what the model needs to do: recognise images, read text or documents, predict demand, or optimise schedules or prices? The selection guide shows which companies offer their own service for each task. Each company’s documented work links to a real case study.
Machine learning is the technique: a model learns patterns from examples, such as how to recognise images or predict demand. Data science is the broader field around it, from analysis to working with your team. This list is about what you want built; for the collaboration side, see the best data science consultancies.
A model ages: the data changes and its predictions gradually become less accurate. That’s why monitoring, retraining and redeployment are part of the job. A few companies on this list handle that themselves. If you’re looking for a specialist provider, see the best MLOps consultancies.
