Case Study: Zurich Switzerland achieves faster fraud investigations and 50K hours saved per year with Neo4j

Faster Fraud Investigation with Neo4j

Zurich Switzerland, part of the Zurich Insurance Group and serving over 1.4 million customers, faced a surge of automatically generated fraud alerts that overwhelmed its 25 field investigators. Rule‑based scoring produced many opaque alerts with little context, forcing investigators to hunt across systems for links between policies, customers, payments and external data, which made triage slow and inefficient.

The insurer deployed Neo4j Enterprise with Linkurious to visualize about 20 million nodes and 35 million relationships, bringing all relevant data into a single graph view and linking new information in real time. The combined rule‑based and graph approach sped investigations (some cases closed within a day), uncovered new fraud patterns, improved investigator productivity and satisfaction, and is saving an estimated 50,000 hours per year while enabling further automation with graph analytics.

Zurich

Paul Kühne

Head of Fraud Prevention