01 The challenge
Submission volumes were rising faster than the underwriting team could grow, and the queue was forcing the business to turn away opportunities. Each case meant pulling data from several external sources and applying rating logic manually, so decisions varied between underwriters and took days.
02 What we did
- Data aggregation from external risk, credit, property and claims sources into a single case view
- Rating and risk models encoding the insurer's appetite, with explainable outputs
- Straight-through processing for cases inside agreed thresholds, with referral rules for everything else
- Underwriter workbench presenting the case, the model's reasoning and the recommended terms
- Portfolio monitoring so loss ratio and appetite drift are visible to management
03 The results
- Standard policies approved in minutes rather than days
- A large majority of applications processed straight through with no increase in loss ratio
- Underwriter time redirected to complex, high-value risks, lifting monthly capacity
- Broker satisfaction improved markedly