
These models deliver enhanced claim scores and reason codes to detect potential fraud and support investigations.
Vendor
Verisk
Company Website

Claim Scoring’s precise predictive analytics swiftly and accurately detect claims fraud so you can fast-track settlements.
How do you detect fraud while processing claims quickly?
Many claims departments lack the IT resources and broad industry data needed to develop effective fraud analytics systems. But implementing third-party analytics can be expensive and time-consuming. Plus, fraud models can have confusing outputs, wasting SIU time.
Fight claims fraud with predictive analytics
Claim Scoring™ lets you score and evaluate claims quickly and effectively using your own real-time data plus aggregated data from 95% of P&C industry carriers.
Fast, accurate fraud detection
Machine learning and predictive models provide accurate claim scores and reason codes for swift triage, evaluation, or settlement decisions.
The world’s largest P&C claims database
Claim Scoring leverages data from over 1.8 billion claims industry-wide to identify fraudulent claims, surpassing models based solely on your own data.
Respond quickly to emerging risks
Build custom fraud scenarios to test ideas using expert business rules and other data points to create more efficient, flexible fraud detection.
Claims fraud detection that stays ahead of fast-evolving schemes
Discover how Claim Scoring uses machine learning, vast data sets, and predictive models to confidently expedite fraud detection and claim triage.
Visual insights to improve claims investigations
Dashboards show you:
- Scored claims in real time
- Deep insights into reason codes used in claim scoring
- Questionable claim exposures in your geographical book of business
- Insights into loss history dynamics via ClaimSearch®
Features and capabilities
Claims predictive analytics through AI and machine learning
Claim Scoring uses multiple variables to identify claims for further investigation and uses fraud scenarios to give SIU meaningful context for model output.
Custom fraud rules for efficient claims fraud detection
The sandbox environment in Claim Scoring lets you create and test custom rules, create targeted alerts for your modeling, and collaborate internally with data scientists to optimize scoring.
Access to more than 1.7 billion claims records
Claim Scoring compares claims to records in the world’s largest P&C claims database and leverages third-party data and civil/criminal records.