I bridge the gap between cutting-edge research and practical implementation by building ML systems that deliver real impact.
ABOUT NIEK
Who am I
I'm Niek Mereu, a Senior Machine Learning Engineer based in Europe, with a passion for turning complex data challenges into elegant, production-ready solutions.
I'm at my best when translating messy, real-world problems into systems that actually work. Not just in theory, not just in notebooks, but under real constraints, real users, and real expectations.
For me, machine learning does not stop at the model. It includes integrating models into backend systems, designing reliable architectures, and ensuring they remain scalable and maintainable over time.
Academic foundation
My journey into data science began at Leiden University, where I earned my Master’s degree in Statistics.
My thesis on heteroskedasticity in online linear regression shaped the way I still approach machine learning today: with a strong respect for uncertainty, assumptions, and the limits of models.
That statistical grounding continues to influence how I design systems end to end. I value not chasing complexity for its own sake, but choosing approaches that are explainable, defensible and fit for purpose. This is both from a modeling perspective and from an engineering standpoint.
BACKGROUND
5+
Years of experience
Building production-grade ML systems across recommendation engines, GenAI and MLOps.
0
Demo-only projects
Everything I build is designed to run in the real world, not just impress in a notebook.
📍
Based in Europe
Available for remote engagements
APPROACH
How I work and think
Over the years, I've learned that the hardest part of machine learning isn't the model, but everything around it.
Understanding the problem well enough to model it properly. Making trade-offs between elegance and practicality. Building systems that teams can maintain, trust, and improve over time without it becoming a liability.
I'm naturally drawn to the spaces between research and engineering, experimentation and production, data and decision-making. Those in-between spaces are where the real complexity lives and where getting it right actually matters.
I enjoy working closely with product, engineering and research teams to turn abstract ideas into tools people can actually rely on. Not just technically sound, but genuinely useful to the people using them every day.
Sharing knowledge is part of that process for me. Teaching, speaking, and explaining ideas forces clarity. Often is clarity where better solutions start.
If you are interested in some of my talks, feel free to have a look
01
Understand the problem first
Before touching any code, I make sure I deeply understand what's actually being asked and whether ML is even the right tool.
02
Keeping it simple
I choose the simplest approach that solves the problem well. Complexity is a liability in production.
03
Own the full lifecycle
From model design to deployment to monitoring, I won’t hand off at the model boundary and walk away.
04
Communicate clearly
I work closely with product and engineering teams. Good collaboration starts with making complex ideas accessible.
05
Build to last
Reliability, scalability and maintainability aren't afterthoughts. They're built in from the start.
MISSION
What drives me
What motivates me most is bridging the gap between cutting-edge research and practical implementation. I believe machine learning should earn its place by being reliable, scalable and most importantly it should be measurable.
If it is able to do so, we can create real impact rather than impressive demos. That distinction matters deeply to me. Anyone can build something that looks good in a presentation. Far fewer can build something that holds up six months later under real load, real data, and real users.
If any of this resonates and if you're tired of mediocre ML holding your product back, I'd love to hear your story!
I don’t build things to impress. I build things to last.
💬
No hard sell. Just a conversation about what you’re building and whether I can help.