Priya's Tuesday, Two Ways

Priya, a commercial lines underwriter at a regional insurer.

Priya arrives Tuesday to three new submissions and a broker email about a bakery renewal. The bakery has delivery vans, a late-night counter, and missing loss runs. In ChatGPT, she uses the Application Intake Review prompt to turn the broker's email and attached application into a one-page summary. It lists the named insured, payroll, vehicle count, and the two items still missing: three years of loss runs and the sprinkler certificate.

Next, she opens Claude and uses the Risk Factor Assessment prompt. She pastes the summary and asks for risk factors grouped by property, liability, and auto, with questions to verify. It flags the fryer, overnight cleaning crew, and a prior slip claim from two years ago. Priya checks each point against the file and adds a note about the bakery's new delivery route.

Later, she uses Gemini for Premium Calculation Support. She asks it to explain the workers compensation rating factors and build a simple worksheet with class codes, payroll, and experience modification. The first pass has the wrong class code for the delivery drivers, so she corrects it and reruns the math. She then uses ChatGPT again for Broker And Client Communication to draft a polite email asking for the loss runs and sprinkler certificate, with a clear deadline.

By 4:30, Priya has reviewed the bakery file, updated her notes, and sent the broker request. She leaves on time, picks up her daughter from soccer practice, and still has energy to cook dinner. The prompts did not make the decision. They cleared the path so she could make it well.