About
To help save time and cost during motor claims processing, Digit built an in-house AI-enabled tool that detects damages to the vehicle based on the pictures provided, analyses and predicts the impact that the damage has created to the vehicle, assess
Innovation presentation
Digit in a short span of 6 years had significantly increased its motor market share and among a total of 25 insurance companies, had become the fourth largest motor insurer in the country. There was a need to make sure our customer service doesn’t get affected, while also making sure we keep a check on our motor loss ratios. This made us realize, there were two major tasks at hand. First, our claims assessment needed to be faster so that the motor claims team does not spend significant time on a single claims case. Second, we needed better understanding of the type of parts that might need replacing or repairing based on the type and impact of damage. This recommendation was typically given first by car workshops after assessing the vehicle and later approved by the surveyor post evaluation.
Uniqueness of the project
Automating this process through AI and predictive analysis could solve two big problems for us. First, the ability to automatically assess vehicle damage would mean the motor claims team would have to spend less time on every case and could issue the Order Form to commence the work faster. Second, the ability to get a machine-based recommendation on the parts that might need replacement based on the damage assessment could help control the vehicle repair cost as it would give us better control over negotiations for vehicle parts repairs and replacement.
· Making claim assessment faster and with better precision.
· Detect vehicle damages quickly and more accurately.
· Control claim payment outgo through better repair recommendation.
· Standardize the vehicle component recommendation based on make, model and vehicle type.
· Arrest the need to unnecessarily replace vehicle components that are not due to accidental damage.
· Mitigate fraud through AI-based image processing for damages instead of solely relying on workshop submissions.