Senior Predictive Modeler – Portfolio Analytics, CUO Portfolio Management
Our global Portfolio Analytics team is seeking an experienced Predictive Modeler to both build and support other team members in developing strong data driven pricing models for all Lines of Business across AGCS. This person will need to collaborate internationally with various stakeholders, both within Portfolio Management and across other AGCS functions. Furthermore, this individual will support Portfolio Analytics in other strategic AGCS projects, such as developing and driving data driven approaches to portfolio management as well as helping to drive forward the overall analytics capabilities of the team.
The Portfolio Analytics team is part of CUO Portfolio Management at AGCS and is responsible for driving the use of analytics within portfolio management across AGCS globally. CUO Portfolio Management is a recently created business function who partner with the LoBs through the provision of in-depth profitability understanding. As well as fostering collaboration and best practice sharing across all the AGCS lines, Portfolio Management provide a cross LoB perspective to the Board to support strategic decision making. The team is spread across London, Munich, Bucharest, Chicago and New York.
- Build predictive pricing models for all Lines of Business at AGCS
- Present and communicate modeling results and recommendations to internal stakeholders
- Act as an internal consultant to the wider Portfolio Management team, providing guidance and training on building data driven pricing models
- Help to improve the overall technical analytics capabilities of the Portfolio Analytics team
- Drive the use of analytics to steer portfolios within AGCS by developing new approaches and successfully them selling to the business
- Develop an expert knowledge of the company’s data and data systems
- Function as a project lead, manage projects and mentor junior modelers
- 4+ years experience in building and implementing predictive models in the insurance industry, preferably for P&C commercial lines.
- Qualified P&C Actuary preferable
- Strong knowledge of P&C insurance and pricing techniques.
- Advanced degree in a quantitative field, preferably with significant emphasis on predictive modeling, data mining or machine learning
- Thorough, hands-on understanding of advanced modeling methods such as regression, GLMs, GAMs, decision trees, ensemble and regularization techniques etc.
- Ability to work efficiently and effectively with large data sets from a variety of sources, including both internal and external data. Strong programming skills in SAS (preferred), SQL, R, Python or equivalent.
- Experience building models on such data sets using statistical software (SAS, SPSS, KXEN, Stata, Emblem, R, Python etc.). Significant experience with SAS and Emblem is highly desirable.
- Demonstrated ability to effectively interpret data and modeling results, and to then communicate them to audiences with differing levels of technical understanding
- Ability to express ideas and messages clearly, both written and verbally
- Proven track record of training up others in predictive modelling related skills and programs such as Emblem
- Skilled in project management
- Organized, self-motivated, results-oriented, and very resourceful