From physics to commodity markets
My background is in theoretical physics — which might seem like an unusual route into commodity trading, but the skills transfer more directly than you might expect. Modelling complex systems, working with large datasets, and building quantitative frameworks are as relevant on a trading desk as they are in academia.
Over almost 20 years, I worked across commodity trading companies and banks in a range of technical and analytical roles. That breadth gave me something I now consider my most valuable asset: I understand both the business and the technology, and I can move between the two without losing anything in translation.
Why I went independent
In 2020 I started working independently, and founded CDMS two years later. The reason was straightforward: I wanted the freedom to focus on the work I find genuinely interesting — building technical solutions for commodity markets, without the constraints of a large organisation.
Commodity markets are complex, data-intensive, and fast-moving. They are an environment where I think I can add real value — because I have seen the problems from the inside, across trading, risk, and research, and I have the technical range to solve them properly. Every business works differently, and the way my clients trade, manage risk, and use data is what drives the architecture — not a standard template.
On AI — from sceptic to builder
I have been very sceptical of AI initially, but decided that I had to have an informed opinion on the topic. What I found is that in the right context, AI can solve problems that nothing else could.
That is what led me to start AIDA with Niels Eweg, my partner for this project. We saw that in complex data environments, it is difficult to access, retrieve and understand data. Traditional systems are built for a specific application, but lack the flexibility to answer new questions. We asked ourselves whether some of that effort could be delegated to an AI. The answer turned out to be more nuanced than expected: while AI works very well with text, working with structured data is a completely different challenge. Solving that problem properly became the start of our AI adventure.
Independent, but not alone
CDMS is a focused, independent operation — which means you always work directly with the person responsible for the outcome. At the same time, I have built a small network of exceptionally skilled collaborators over the years: specialists I have worked with before and whose work I can stand behind.
For larger projects, or where a specific expertise is needed, I bring in the right people. The team scales with the project. The relationship and the accountability stay constant throughout.