Risk and credit
Fraud detection, credit risk and operational risk models
Decisions that must be fast, explainable and right most of the time: automated credit decisions, fraud scoring on retail transactions, and supplier risk mapped as a network rather than a scorecard.
What this usually looks like
- Credit decisions take days because a person re-reads every file.
- Fraud rules block good customers and let the new pattern through.
- Supplier risk is a scorecard nobody opens until a supplier fails.
- The model cannot be explained to the regulator, so it cannot be deployed.
What you get
Automated decisioning with a human lane
Clear approvals and refusals handled by the model, the ambiguous middle routed to a person, thresholds set on cost rather than on accuracy alone.
Fraud scoring that adapts
Transaction and behavioural signals, retrained on a cadence, with the false-positive cost on the dashboard next to the catch rate.
Risk as a network
Which suppliers, accounts or nodes the operation cannot lose, found before the disruption rather than after.
The handover
An explanation for every decision, monitoring for drift, and documentation the compliance team can read.
Selected work
Marketing Science
Automated credit decisioning system
Machine learning-powered credit decisioning system leveraging alternative data from social media and digital footprints combined with traditional credit data to improve risk assessment accuracy and speed.
Decisions 10x faster, operating cost ↓60%, default prediction +25%
Risk & Operations
Supplier risk intelligence framework
Network-based supplier risk modeling that identifies systemic vulnerabilities and "too-central-to-fail" suppliers before disruptions occur.
Too-central-to-fail suppliers identified before disruption, prioritised by systemic impact
How it starts
Diagnostic first, build second.
01 · Decision System Diagnostic · From €12,000
Three months, one critical decision. We establish which decision the model must make on its own, measure the cost of each error, test what the data can support, and design the review cadence. You get evidence, a target architecture, and a costed roadmap.
02 · Design and build
Scoring models, the decision rules around them, the human lane, and the pipeline that runs it. Scoped from the roadmap, priced on the engagement.
03 · Operate and hand over
Monitoring, drift controls, documentation, and the training to run it without me.