Expertise
Always in the middle of it, never on the sidelines: thinking in procedures, having seen countless fraud patterns, markets built, systems repaired, workflows improved and teams led. The complete package, in a large group as much as in a mid-sized business.
Core competencies
Fraud prevention across the whole path: from assessing the risks through the rules and procedures to the daily handling of cases. Rule sets of every kind, from fixed rules through variable thresholds to machine learning models, and the layering that turns individual checks into a net. Plus fast help with live incidents and with scams that are newly appearing.
International payments, steering the payment method mix down to the individual country, conversion and cost optimisation, buy now pay later, business rules, credit management, credit checks as well as dunning and collections.
Anti-money-laundering and transaction monitoring, implementing regulatory requirements such as GDPR, the German Money Laundering Act and BaFin expectations, as well as data protection as a certified data protection officer.
Organisational design and role concepts, transformation support, KPI frameworks across the entire customer journey, data-driven performance management as well as project leadership, interim management and team development.
Artificial intelligence
I have been working with it since it went by a different name in fraud detection. Part of that is knowing what is actually on offer in the market, what the tools can do and where their limits are.
Career
It began with B2B credit management: steering receivables worldwide, granting limits, acting on late payment. Put simply, it was about the money that did not come back. Fraud prevention grew out of that, because the two questions behind it cannot be separated cleanly: who cannot pay, and who never intended to? The same data, two answers, two completely different countermeasures. Confuse them and you send reminders to offenders while treating customers like offenders.
Then came years in international online retail, where this turned into a craft of its own: check routes built for new country markets, rule sets of every kind designed and readjusted again and again, fraud systems consolidated across country entities, and the first in-house machine learning models built together with IT. Fixed rules, variable thresholds and learning models are not alternatives to one another but layers that sit on top of each other and have to catch what the others miss. And none of them holds by itself: a rule that convinces in a test run says nothing about how it behaves after three months in live operation. Readjusting is not rework, it is the operation itself.
Alongside that, deputy data protection officer, today certified. That was no sideshow: the very data from which fraud can be detected is the data most in need of protection. Build detection without thinking about that and you build yourself the next problem.
After that, responsibility for payments and fraud in international retail, including an in-house buy now pay later solution for the marketplace: payment method steering, conversion and risk as one shared calculation instead of three separate goals. Most recently a division of up to 58 people, bringing credit checks, limits, fraud prevention, money laundering and communication with authorities under one roof. Only at that breadth does it become visible how much the areas depend on each other. The stations along the way were bonprix, DOUGLAS and Otto Payments, as well as various startups.
Independent since September 2025. For the first time in a position to take this experience wherever it is needed, instead of keeping it inside one house.
BackgroundA working life in fraud and payment functions, from hands-on project work to running a division. Plus the daily struggle with systems that did not do what they were supposed to.
RelevancePlan at the drawing board and you build procedures that do not protect against fraud. Controls fail at the exception, not at the concept.
BenefitI know which control survives daily business and which one gets worked around after four weeks.
BackgroundDeveloped training frameworks for entire departments, plus further education in business psychology and behavioural economics.
RelevanceScams use time pressure, authority and the reluctance to ask. The same forces decide whether a control is actually lived.
BenefitTraining after which a scam gets recognised, rather than a leaflet signed.
BackgroundOn conference stages from Hamburg to New York since 2018, plus podcasts. Before that a blog of my own on internet law. Over the years that has grown into a network reaching across industries and houses.
RelevanceA new scam appears in no situation report before it has done damage.
BenefitIn an acute case I know within hours whether others are seeing the same thing.
BackgroundA good dozen fraud and credit bureau solutions in productive use, plus benchmarks and an international RFI.
RelevanceAttack methods go out of date within months. Since AI entered the picture, even faster.
BenefitI can tell a vendor promise from an actual capability.
Projects
Across industries and roles, from the first analysis through into live operations.
Whether it is a vendor, a build from scratch, an organisation that is not running smoothly, or an acute case: tell me what it is about and you will get an honest assessment.
Get in touch