Underwriting risk in a data-rich environment: Turning healthcare data into underwriting alpha

Healthcare claims data is the richest risk signal in insurance, and most of the industry uses it only to pay claims. Indigo used it to price risk: procedures performed, prescriptions written, patient populations served – 1,500 data points per physician, pulled from a single ID number instead of a 300-question application. The line is physician liability, but the underwriting question is universal: which data actually predicts loss, and which just makes the application look modern? Why do horizontal "AI underwriting" layers stall at a fraction of the job? And how do you grow fast without pricing below your loss ratio? The focus is practical: what health underwriters can take from a carrier that built its edge on their own data.