We are seeking a highly motivated Data Science Intern who is interested in exploring data, experimenting with modeling approaches, and developing analytical solutions that influence business decisions. You will support the Advanced Analytics team by performing exploratory data analysis (EDA), data engineering, feature engineering, and statistical modeling to identify, prioritize, and explore predictive features for insurance loss modeling. The role will involve collaboration with stakeholders from the underwriting, pricing, and operations departments to ensure that data products align with business needs.
Assess the business value of a third-party data source using an established scoring rubric
Learn core insurance concepts and build foundational knowledge of Property & Casualty insurance products, markets, and underwriting processes
Conduct exploratory data analysis (EDA) on insurance and oil & gas datasets to understand patterns, data quality/coverage, and modeling opportunities
Explore data engineering approaches and test methods for entity matching, including geospatial and temporal linkage techniques
Engineer features by creating, validating, and maintaining derived variables that support analytical and predictive modeling
Collaborate with business and technology partners to clarify requirements, understand key variables, and align on acceptance criteria
Prototype predictive models and compare lightweight models against existing benchmarks
Quantify value by estimating cost/benefit and return on investment for proposed features, models, and data products
Develop analytics workflows and implement repeatable processes for EDA, feature creation, and model evaluation
Plan model integration by working with engineering teams to outline how analytical models will be incorporated into production systems and reporting tools
Create analytical datasets and dashboards to prepare datasets and build reports that help stakeholders make informed decisions
Ensure data and model quality by conducting QA checks on datasets, metrics, and prototype models
Maintain documentation for datasets, features, models, and analytical processes
Automate repetitive tasks using scripting and AI tools to streamline work and improve productivity
Previous quantitative research experience (statistics, applied mathematics, computer science, or related field) in a course, personal project, or previous role
Experience using databases and data modeling concepts, with proficiency in SQL (cloud data platforms a plus)
Founded in 1967, W. R. Berkley Corporation is an insurance holding company that is among the largest commercial lines writers in the United States and operates worldwide in two segments of the property casualty insurance business: Insurance and Reinsurance. Each of the operating units within Berkley participates in a niche market requiring specialized knowledge about a territory or product. Our competitive advantage lies in our long-term strategy of decentralized operations, allowing each of our units to identify and respond quickly and effectively to changing market conditions and local customer needs. This decentralized structure provides financial accountability and incentives to local management and enables us to attract and retain the highest caliber professionals. We have the expertise and resources to utilize our strengths in the present environment, and the flexibility to anticipate, innovate and respond to whatever opportunities and challenges the future may hold.
