Jobs

WHAT IS THE APPLIED RESEARCH & DATA ORGANIZATION AT THE HARTFORD?

The Hartford's Property & Casualty organization formed its Applied Research & Data organization in June 2006 to set forth an innovative and market-leading approach to cutting edge quantitative research within the insurance industry.

The organizational design focused on the optimization of the value a Research & Data team brings to the business units it supports. This group operates within a culture of shared philosophies, team work and an appreciation for ideas and thought leadership.

A cornerstone of our Research and Data organization is its unique structure which is designed as a unified research team leading research initiatives across the enterprise in product, pricing, and claims. This team reports to a single Senior Leader.

Today, this team of high-powered and highly skilled research professionals supports over $10 billion/year in personal, small commercial, middle market and specialty lines premium. Our Researchers have the opportunity to work on many different types of highly visible and well supported projects throughout the P&C organization.

If you are an experienced quantitative researcher with an interest in joining a dynamic and highly collaborative research organization, where the sharing of information and ideas across people and project teams is customary, we'd like to speak with you!

WHAT ARE WE LOOKING FOR?

We are looking for a Sr. Quantitative research scientist responsible for carrying out the research and prototyping tasks for the Ideation & Prototyping group within the Applied Research & Product Development (ARPD) organization. The Ideation & Prototyping group is a forward-thinking area dedicated to researching and enhancing state-of-the-art advances in data-related processes, practices, and analytics within (and across) P&C product lines.

The quantitative research scientist performs research, conducts exploratory data analysis, builds predictive models, develops simulations, etc. and instantiates them as prototypes.

Critically important is an understanding of the different data-demands and requirements in data mining vs. traditional statistical analysis. Modeling tasks can span Property & Casualty Personal and Commercial data, internal and external data sources, and be based on massively large datasets. Both traditional statistical analysis (regression, GLMs) and data-mining techniques (trees, NNs, etc.) are in scope.

In collaboration with data research analysts, business analysts, and solution developers, the individual understands the scope and analytical objectives, develops appropriate models and presents results within project timelines and deadlines. THIS POSITION IS LOCATED IN CAMBRIDGE MA.

Further responsibilities and accountabilities include the following:

• Modeling
• Lead and guide the data-preparation aspects that precede model development.
• Develop and prototype models employing advanced statistical and/or data-mining techniques as appropriate (regression, GLMs, clustering, decision-trees, neural networks, etc.)
• Develop new application specific techniques as appropriate (e.g., fraud detection, price optimization)
• Recommend and initiate model redevelopment, and enhancements beyond prototypes as required.
• Adopt the use of data-standards related to quantitative model building (e.g. PMML, XML, etc.) to facilitate transition of prototypes into implementation and deployment.