Korva is a scroll of the internet’s best ideas.
From fraud investigator to data scientist
Fraud investigations
At the Florida regulator of banks and finance, Todd Combs worked alongside the SEC and the FBI. One of his last cases was a huge insurance fraud in which several federal judges were involved and went to prison, and he unraveled it.
The math exam
Friends of his worked at Progressive and loved their jobs, which struck him as odd for an insurance company. He applied. Most applicants were headed for claims jobs and were given a math exam, and his result had Progressive tell him to go over here instead, to what would today be called data science. The term hardly existed back then.
Building industry tools
He had one small state, and from there one thing led to another. He ended up rolling out credit and telematics for the industry, and he has his name on patents.
Why is insurance fraud so costly?
The Coalition Against Insurance Fraud, an industry and consumer group, estimated in 2022 that fraud costs Americans about $308.6 billion a year across all types of insurance. That covers everything from staged auto accidents to fraudulent health billing and fake life insurance claims. Because of that cost, most states run dedicated fraud bureaus that investigate suspected schemes and refer cases to prosecutors. The FBI also investigates insurance fraud, including cases that cross state lines.
How did data scientist become a job title?
DJ Patil and Jeff Hammerbacher, who led data teams at LinkedIn and Facebook, are credited with popularizing the title around 2008. Before that, similar work went by names such as statistician, analyst, or actuary. Harvard Business Review called data scientist the sexiest job of the 21st century in a 2012 article by Thomas Davenport and Patil.
How does telematics change car insurance pricing?
Telematics-based insurance uses sensors or smartphone apps to record driving behavior such as hard braking, speed, and time of day. Insurers feed these measurements into pricing models to charge according to how a person drives rather than group averages. The approach is called usage-based insurance.