Scores that make you go hmm: When predictive models and underwriters don’t see eye to eye

Predictive models like Irix Risk Score are insurtech marvels that help life carriers quickly issue more policies, improve customer service, and add distribution channels and product offerings. Most of the time, underwriters and Risk Score coexist in peace and harmony. But sometimes a score just seems off.

If those “unintuitive results” are dogging your underwriting team, this webinar (presented with Munich Re) could be the “Eureka!” moment you need to make peace with scores that don’t jibe with traditional methods. In a case-by-case analysis, we’ll compare human underwriting decisions with scores and see where they agree and where they part ways—and why. You’ll get insights into how the Risk Score model works and how human art and data science can come together for better-than-ever risk assessment.

Topics covered

• Strengths and weaknesses of predictive modeling and traditional underwriting methods, plus the differences between them
• How Risk Score interprets applicant data and issues scores based on layers of inputs
• Rationale for human underwriting decisions using traditional methods
• Case-level, side-by-side comparison of traditional results and scores

Who should attend

Underwriters, actuaries, medical directors, and other clinical and risk professionals at life insurance carriers will benefit from a more mature understanding of predictive modeling and risk scoring in life underwriting.

Michael Niemerg, FSA, MAAA

Principal and Director, Data Science, Milliman IntelliScript

Rachel Eberle

Director, Underwriting Risk, Munich Re Life US

Sue Bartholf, FSA, MAAA

Director, Solutions Consulting, Milliman IntelliScript