Thanks to the collaboration between the experience of Msa and Kube Partners, a company specialized in creating advanced technological platforms for the insurance sector, we made the anti-fraud engine Detector.

Overall Fraud Index

The combination of three different elements


Intelligent Methodology of analysis

Detector represents the new frontier of software for the research and management of insurance fraud. The web-based application uses classification and prediction algorithms to check the policy and customer portfolio, analyzing large amounts of data searching for recurring patterns. In terms of functionality and usability, this tool offers technically advanced calculation methods and can be easily integrated into the company’s processes. Detector performs a triple function: it automatically identifies the fraud index; it is the tool for managing the investigation activity; it calculates the savings on the overall cost of the claim.

How it works

How it works


It highlights the summary of the liquidators’ experience

Detector converts the procedure followed by an expert settlement agent into a sequence of automated queries sent to the databases (claims, portfolio, registry, accounting). A structured checklist and query of the Ivass claims database will guide the liquidators’ identification of anomalies.


Each claim and the parties involved are included and assessed based on the territorial context.

The analysis based on a database developed by Kube Partners (“Risk Map”) collects external data detailing them at a territorial level, according to socio-economic factors and the insurance nature. Detector geographically places the parties involved and the places of the accident to highlight possible suspicious relationships and actions. The system also makes use of integration with Google Geolocation and Google Maps Image API.


Identify hidden relationships and possible fraudulent networks.

Identifying links between the parties, events, and critical characteristics makes it possible to detect previously unidentified networks and connections and highlight possible fraudulent activities. Detector graphically shows the relationships identified and provides details with a high degree of detail.


Independent assessments using artificial intelligence systems.

Detector uses multiple inference engines, including Google Prediction API and Neural Networks, based on non-linear classification algorithms and behavior prediction. These provide a reliable assessment of potential fraud by comparing the claim with the history of other fraudulent events and comparing different methodologies.


Visit the dedicated section for more details

©Msa Multi Serass

Msa – Multi Serass S.r.l, a limited liability company, Subject to the Direction and Coordination of MSA Mizar SPA

Registered office: Via Sangro, 15 – 20132 Milan – R.E.A. MI – 2506361 – Fiscal Code / VAT number 10116240960