IBM Netherlands: We don’t believe in one-size-fits-all—specialized AI is the future
InterviewsSaturday, 13 September 2025 at 11:07
AI Wereld visits IBM Netherlands in Amsterdam, welcomed at the IBM Innovation Studio. AI is reshaping business at breakneck speed. But does ChatGPT actually work on the job? IBM says no—and explains why.
We speak with Sophie Kuijt, CTO for Northern and Central Europe, and Damiaan Zwietering, Developer Advocate Data Scientist at IBM. Benjamin Timmermans, showcase experience lead at the Innovation Studio, also gives us a tour of the IBM Experience.
How is IBM’s AI approach different from popular consumer models like ChatGPT?
Damiaan Zwietering (DZ): “We don’t believe in one-size-fits-all. Big, generic models are often inefficient and lack domain depth. We build smaller, targeted language models—finance, government, industry—trained on trusted datasets we fully understand. They’re more energy-efficient and easier to explain.”
Damiaan Zwietering (DZ): “We don’t believe in one-size-fits-all. Big, generic models are often inefficient and lack domain depth. We build smaller, targeted language models—finance, government, industry—trained on trusted datasets we fully understand. They’re more energy-efficient and easier to explain.”
How does IBM ensure AI is safe and accountable?
DZ: “At IBM we often say: we are The Governance Company. Explainability, monitoring, and compliance are built in from day one. Organizations can instantly see if their AI still meets the EU AI Act. If not, you get a clear alert.”
DZ: “At IBM we often say: we are The Governance Company. Explainability, monitoring, and compliance are built in from day one. Organizations can instantly see if their AI still meets the EU AI Act. If not, you get a clear alert.”
Sophie Kuijt (SK): “Governance isn’t an afterthought—it’s part of the architecture. Our tech is scalable and anchored in clear principles derived from IBM’s core values.”
Where is IBM already delivering AI that works?
SK: “A great example is an HR chatbot we built with SLTN. Inside IBM, we rolled out AskHR, a conversational AI agent. AskHR already answers 94% of all HR questions automatically. From leave balances to changes—it all runs through one intuitive interface. That’s cut turnaround times and improved the user experience.”
SK: “A great example is an HR chatbot we built with SLTN. Inside IBM, we rolled out AskHR, a conversational AI agent. AskHR already answers 94% of all HR questions automatically. From leave balances to changes—it all runs through one intuitive interface. That’s cut turnaround times and improved the user experience.”
DZ: “What started as a simple assistant has evolved into an agent that takes action, like booking vacation days. It operates in a safe, controlled environment—vital when handling sensitive employee data.”
Where does IBM fit in the broader AI ecosystem?
SK: “IBM has decades of foundational AI research. Much of our innovation is open source; via Hugging Face you can download our Granite models and our fine-tuning tool InstructLab. At the same time, we deliver full enterprise solutions via platforms like IBM Cloud, AWS, and Azure.”
SK: “IBM has decades of foundational AI research. Much of our innovation is open source; via Hugging Face you can download our Granite models and our fine-tuning tool InstructLab. At the same time, we deliver full enterprise solutions via platforms like IBM Cloud, AWS, and Azure.”
How central is open source to your strategy?
SK: “Essential. Open source brings transparency—not just what the model does, but how it’s built. We share models and the tools to adapt them, giving others the same capabilities we use internally.”
SK: “Essential. Open source brings transparency—not just what the model does, but how it’s built. We share models and the tools to adapt them, giving others the same capabilities we use internally.”
Which customers are leading in AI adoption?
SK: “Heineken is a strong example. They use AI end-to-end: from sales and marketing to the breweries. They’re unifying insights from local datasets and pushing hard on productivity gains with IBM as their tech partner.”
SK: “Heineken is a strong example. They use AI end-to-end: from sales and marketing to the breweries. They’re unifying insights from local datasets and pushing hard on productivity gains with IBM as their tech partner.”
DZ: “We also see a clear shift in finance, insurance, and government. First, AI makes processes smarter—what we call ‘AI+’. Eventually, that evolves into AI running processes end-to-end.”
IBM levert ook technologie aan de Formule 1 app van Ferrari
How does IBM help customers adopt AI?
SK: “We believe in co-creation. We run hackathons and give all our consultants access to the IBM Consulting Advantage platform. That’s where practical ideas emerge—and they’re scalable internally, which speeds up learning.”
SK: “We believe in co-creation. We run hackathons and give all our consultants access to the IBM Consulting Advantage platform. That’s where practical ideas emerge—and they’re scalable internally, which speeds up learning.”
DZ: “For example, a client wanted to ‘chat’ with their documentation. Six months ago, we built a custom solution. Now that same client can upload PDFs in our cloud and start using an agent immediately.”
What about costs and technical barriers?
DZ: “Lower than you think. For simple use cases like a document assistant, you’re often under a thousand euros a month. And we only monitor resource usage for billing, not data content—unless a client explicitly opts in to improve the model.”
DZ: “Lower than you think. For simple use cases like a document assistant, you’re often under a thousand euros a month. And we only monitor resource usage for billing, not data content—unless a client explicitly opts in to improve the model.”
SK “And choosing IBM means guidance—from MVP to scalable enterprise deployment. We help with monitoring, aftercare, and cost control, always aligned with the client’s goals.”
How does IBM attract new AI talent?
DZ: “We’re active locally with innovation centers in Groningen and Eindhoven, partner with universities and applied sciences, and give guest lectures. The mix of consulting, research, and technology in one company appeals to young talent.”
DZ: “We’re active locally with innovation centers in Groningen and Eindhoven, partner with universities and applied sciences, and give guest lectures. The mix of consulting, research, and technology in one company appeals to young talent.”
SK: “Students often discover during their thesis or internship how broad and deep IBM’s tech is. That makes sense—we usually operate under the hood. We’re there—say, in every financial transaction—but you don’t see us.”
Finally: IBM’s view on AI’s future and AGI?
DZ: “We’re not chasing generic AGI. Our focus is targeted, organizational AI. There’s massive value to unlock—responsibly, at scale, and with transparency.”
DZ: “We’re not chasing generic AGI. Our focus is targeted, organizational AI. There’s massive value to unlock—responsibly, at scale, and with transparency.”
SK: “At IBM, we introduce technology in a controlled way. AI must work—and be trustworthy—at scale. That demands governance from day one.”
IBM isn’t buying the hype—it’s betting on real-world impact. If IBM has its way, AI stays reliable, explainable, and workable. Whether that holds up remains to be seen.
Loading