Probably the tip of the iceberg. BUT- Before jumping into AI and machine learning look at how much data is in silos or just not harvested. Probably between 80% and 90% which males AI a non-starter.

Simply digitising the claims process by allowing claimants to complete customised templates capturing essential information- different for vehicles, buildings and pets for example.

See "Digital Tools and Power to Customers".

These allow claims handlers, via SMS & email , to  send links requesting claimants to use phone & tablet to submit photo and video evidence to validate the claim. This self-service is proven to result in up to 30% walkaways with an immediate impact on insurer's bottom line.

Then being able to analyse ALL the data from a claims notification form (CNF) means that key fraud indicator (KFI) rules can be applied to help identify further fraud. These are tools fraud investigators can use to combine intuition and "a nose for fraud" with analytics- augmented intelligence. 

See "Augmented Intelligence rather than AI"

Last but not least, being able to access and analyse data from new digital platforms AND legacy systems of record opens up the opportunity to apply these KFI rules to hunt out the seasoned, serial fraudster. 

Apply this digital transformation today whilst planning for the future data management and AI/machine-learning tools and capabilities to automate simpler claims and products. 

But don't run before you can walk. Don't spend millions when you can adopt the processes above on a reasonable cost per claim processed.

See how at "See every claim, see every risk"