Our client
Our client is a leading Data-as-a-Service platform provider in the Benelux, specialising in product and article information across hardware, tools, construction materials, paint, and automotive. They enable suppliers, procurement organisations, and retailers to centrally manage, standardise, and optimise product data at scale.
Raw supplier data — often messy, multilingual, and inconsistent — is transformed into high-quality, structured product information through their platform. AI is not a supporting feature here; it is the core mechanism driving data quality, efficiency, and scalability. The organisation is now building the foundations of its next-generation AI-native platform, and this role is central to that work.
The role
As AI Architect, you will combine scientific rigour with real engineering ownership. This is a dual-track position: you will be hands-on — building, fine-tuning, and evaluating models in production — while also shaping the long-term AI strategy and advising leadership on where the organisation should invest next.
You will be the internal authority on AI: the person who sets the scientific standard, challenges assumptions, and ensures that what gets built actually works reliably at scale. Working closely with data, engineering, and product teams, you will own the AI architecture from research through to deployment.
Tasks & responsibilities
- Design and implement AI models for classifying, normalising, translating, and semantically enriching large volumes of multilingual product data — with rigorous evaluation and production-grade quality standards;
- Evaluate, adapt, and where necessary fine-tune large language models (open-source and proprietary) for structured data tasks; benchmark approaches systematically and document what works and why;
- Architect end-to-end AI pipelines that integrate cleanly with the broader data platform — modular, observable, and built to last;
- Develop validation mechanisms and governance frameworks — including uncertainty quantification, confidence thresholds, human-in-the-loop review flows, and automated monitoring — to ensure reliable and trustworthy AI output;
- Act as a strategic advisor to management on AI-related topics: ground investment decisions in scientific reality, define the organisation’s model selection and evaluation principles, and contribute to the long-term AI vision;
- Drive internal AI adoption by coaching and upskilling colleagues — raising the technical floor of the team without creating dependency on yourself.
Your profile
- A strong foundation in machine learning and NLP — from a research MSc, PhD, or equivalent depth of applied experience; you understand model internals, not just APIs;
- Proven experience developing and deploying AI and machine learning solutions that operate under real quality, reliability, and governance requirements — not just proof-of-concepts;
- Hands-on proficiency in Python and the modern ML ecosystem: PyTorch, HuggingFace Transformers, LLM APIs, and open-source model tooling;
- Solid understanding of data structures, data quality, and ethical AI practices; you approach model evaluation with the same scepticism you’d apply to any scientific experiment;
- Experience within a product or data-platform organisation is a strong advantage;
- Able to communicate technical depth clearly to non-technical stakeholders — without losing the nuance that matters;
- Proactive, independent, and intellectually curious — you follow the research landscape, form opinions about it, and share them actively across teams;
- Fluency in Dutch is a must.
