Remote damage assessment model to reduce frauds and speed up claims for Asia's top Insurer
Client
The client is a top-tier insurance provider in Southeast Asia, offering automobile and personal insurance services across multiple countries, with a large base of policyholders and claims volume.
Project Context
The client wanted to speed up claims and reduce fraud while settling insurance claims. They wanted a mechanism to remotely assess the vehicle damage and send notifications to the inspecting officer.
Challenges
The client provides insurance to automobiles and wanted to speed up claims and reduce frauds.
It was impossible for the inspector to assess thousands of vehicles and assign the correct insurance value, they wanted to automate this process by developing a remote inspecting model and thereby approving claims.
Solution
SquareShift built a real-time video and image processing model to detect the car make and severity of damage.
They built dataset with images of car and automated notification to inspector on the damage details and claim payouts.
Project Objectives
Eliminate physical evaluation for minor accidents.
Accurately match the estimated claim with the final claim amount.
Send notification and expedite claims.
Solution Delivery
SquareShift built a ML model with Car Detection Model, Car View & Damage Classification, Car Severity Detection and Car Damage Estimation.
Claims data model, make and damaged parts were matched against the amount claimed.
Inspectors get a notification of the damaged parts in the image and they can validate with actual damage.
Testimonial
SquareShift’s remote damage assessment solution helped us fast-track claims with greater accuracy, and drastically reduced our fraud rates.
