About the Team:

Data scientists at Ledger are critical to the success of the company. Our team specializes in modeling the behavior of casualty insurance risk portfolios. We analyze and predict the performance of every insurance program that seeks capital through us, and our analyses are critical in determining on what terms, if any, our investors are willing to offer capital for each program. In short, our math forms the ground truth that serves as the basis of negotiations for transactions with tens of millions of dollars at stake.

About the Position:

As our company continues to grow, we are focused on opportunities to use technology to scale efficiently. From a data science perspective, one such core opportunity is providing a seamless infrastructure layer that allows everyone in the organization to easily train, backtest, and deploy statistical models. We heavily rely on Stan for our modeling, which has not gotten as much attention as some other tools in the data science and machine learning communities for this sort of infrastructure.

A complete analytical process for a single insurance program may consist of a half a dozen or more independent statistical models. The identity, configuration and method of combination of these models varies considerably from program to program, and we are responsible for providing daily or weekly performance updates on dozens of these programs. Striking the right balance between flexibility and consistency for deploying these analytical processes to production is another key problem.

If these problems sound fun and interesting, we'd love to have you on our team.

About You:

Successful candidates will have all of the following attributes:

• Strong knowledge of statistics, machine learning, and data science.
• Extensive experience with Python and its associated ecosystem for machine learning (e.g., NumPy, SciPy, Pandas, Jupyter, etc).
• Extensive experience with virtualization, cloud environments, cluster computing, and other techniques for efficiently deploying models for large-scale production use.
• Experience with relational databases and strong ability to write queries in SQL.
• Familiarity with version control, especially Git/GitHub.
• Ability to work independently and communicate ideas effectively.

The following attributes will help you stand out from the crowd:

• A track record of building or contributing to open-source projects.
• Working knowledge of property & casualty insurance, particularly from an actuarial and/or underwriting perspective.
• Experience with Stan or another open-source tool for estimating Bayesian models via HMC.