FinTech – transaction risk management

In digital payments, seconds matter. The faster and more convenient payments become, the more important it is to reliably catch the ones that are not right. That is why we designed and developed a scalable solution for detecting risky transactions, which evaluates them in real time and so increases payment security.
At its core is a weighted scoring model that combines several inputs. The solution evaluates transactions asynchronously through a message queue and passes risky cases on for manual review. We tuned the weights so that potential threats are detected reliably. The solution is running in production.
We designed it with the future in mind. The platform is built so that new services can be plugged in easily and so that it can grow with the client’s needs, providing a reliable foundation for modern fintech solutions.
Beyond that, we introduced single sign-on through OpenID Connect across the internal tools and set up monitoring, alerting and anomaly detection in Prometheus and Grafana.
Technologies:
Backend: Spring Boot, Kotlin
Data & messaging: RabbitMQ, Redis
Infrastructure: Kubernetes (K8s), Dev Containers
Security & monitoring: OpenID Connect (OIDC), Prometheus, Grafana