Data Scientist: Underwriting Analytics & Predictive Modeling

Insight Global; Conshohocken (Montgomery County); Vor Ort; USD 110.000 - 140.000; Vollzeit

Zusammenfassung

Insight Global is seeking a Data Scientist for a full-time role in the Conshohocken, PA area to partner with the Underwriting team and deliver predictive analytics across the department. You will build models distributing actuarial loss costs across risk characteristics and design underwriting levers to hit target loss ratios while staying competitive.

You will apply modern data science techniques (predictive modeling, GLMs, gradient boosting, clustering) to create underwriting models and

Qualifikationen

• Strong grounding in statistics, actuarial science, and underwriting fundamentals.
• Experience in data science or actuarial-adjacent roles; 2+ actuarial exams or equivalent pricing credential.
• P&C insurance experience; commercial lines or MGA/program business preferred.

Aufgaben

• Build predictive underwriting models to hit target loss ratios while remaining competitive.
• Collaborate with Underwriting to socialize findings and influence guidelines.
• Apply GLMs, gradient boosting, clustering to uncover segmentation opportunities.
• Partner with Underwriting leaders to implement lever changes.

Kenntnisse

Mathematical statistics

Actuarial science

Underwriting fundamentals

Data science

Actuarial background

Pricing credentials

P&C insurance experience

Commercial lines

MGA/program business

SQL

Python

GLM

Predictive modeling

AI tools

Tools

Git

Cloud platforms

Power BI

Tableau

Jupyter

Jobbeschreibung

Insight Global is seeking a Data Scientist for a full-time role in the Conshohocken, PA area to partner with the Underwriting team and deliver predictive analytics across the department. You will build models distributing actuarial loss costs across risk characteristics and design underwriting levers to hit target loss ratios while staying competitive.

You will apply modern data science techniques (predictive modeling, GLMs, gradient boosting, clustering) to create underwriting models and