CAR INSURANCE CLAIMS PROVIDER
BUSINESS PROBLEM
This company needed to identify highly probable fraudulent claims in their claims processing system.
SCALESOLOGY IN ACTION
Car insurance claims were analyzed using machine learning algorithms for the following:
To predict whether auto parts should be repaired or replaced.
To identify links between car parts commonly damaged together and to categorize damages based on vehicle type
To identify fraudulent claims where repair shops added atypical or suspicious items/parts to insurance claims
RESULT
An automated process to detect fraudulent claims and a new product created to allow their customers to maximize efficiency when estimating damage on cars.
SERVICE REFERENCES
Comments