Senior Director I, Modeling Data Pipelines

Liberty Mutual•Seattle, WA

About The Position

Leads the development and delivery of end-to-end data pipelines and feature store capabilities that provide data scientists with modeling data at the speed required by P&L vertical teams. Provides strategic and tactical direction for the department and helps shape how MDS, data scientists, and deployment teams across US Data Science (USDS) work with modeling data. The role requires demonstrated ability to lead large-scale change across multiple teams and is responsible for managing, providing technical oversight for, and developing a team of data analysts, MLOps specialists, and subject matter experts. Owns the strategy and execution for modernizing and standardizing the modeling data lifecycle, from data ingestion through model deployment, in partnership with P&L vertical data science, modeling data, and deployment teams. Uses modern pipeline frameworks and scalable architecture patterns to track data lineage throughout feature development. Leads the design and adoption of a feature store strategy that enables features to be defined once, managed consistently across model development, calibration, and deployment, and added to directly by data scientists. Leverages pipeline modernization to develop AI solutions that accelerate feature development, testing, and maintenance. Applies deep knowledge of modern data pipelines and feature management, including tools such as dbt, Snowflake, and Databricks, to pricing model development, calibration, and deployment. Collaborates with USDS, Data Office, DAE, ED&DS, and Product Delivery. The role also provides strategic and tactical direction at the highest levels, owns a relationship with a defined business segment, and is accountable for the performance and results of one or more large, strategically important business units in an extremely large market. It leads complex, cross-functional projects and predictive analytics assignments, shapes the data science community, and leads cross-functional working groups. The position may oversee subordinate managers and coaches the team to understand organizational vision and goals. The employer describes a workplace that values inclusion, supports employees through benefits, workplace flexibility, and professional development, and offers opportunities through Employee Resource Groups. Liberty Mutual is an equal opportunity employer and does not tolerate discrimination on protected grounds.

Requirements

Typically 8 or more years of relevant experience, including leadership experience in data science, analytics, data operations, or MLOps.
Experience sourcing, integrating, and governing data from enterprise data warehouses and external vendors.
Ability to partner with data science, technology, and business stakeholders.
Ability to communicate clearly with technical and non-technical audiences.
Deep knowledge of modern data pipelines and feature management practices.
Experience applying data pipeline and feature management practices to pricing model development, calibration, and deployment.
Ability to lead large-scale change across multiple teams.
Advanced degree in Statistics, Mathematics, Data Science, Computer Science, or a related field preferred.

Nice To Haves

Advanced degree in Statistics, Mathematics, Data Science, Computer Science, or a related field.

Responsibilities

Lead development and delivery of end-to-end data pipelines and feature store capabilities for modeling data.
Provide strategic and tactical direction for the department and help shape how data science and deployment teams work with modeling data.
Manage, provide technical oversight for, and develop a team of data analysts, MLOps specialists, and subject matter experts; the team may include subordinate managers.
Own strategy and execution for modernizing and standardizing the modeling data lifecycle from data ingestion through model deployment.
Partner with P&L vertical data science, modeling data, and deployment teams to modernize the modeling data lifecycle.
Use modern pipeline frameworks and scalable architecture patterns to track data lineage throughout feature development.
Lead the design and adoption of a feature store strategy that supports consistent feature definition and management across model development, calibration, and deployment.
Enable data scientists to add features directly to the feature store.
Use pipeline modernization to develop AI solutions that accelerate feature building, testing, and maintenance.
Collaborate with USDS, Data Office, DAE, ED&DS, and Product Delivery to develop solutions.
Coach the team to understand organizational vision and goals.
Collaborate with business partners to develop modeling tools, dashboards, and decision aids for strategic decision-making.
Own relationships with a defined business segment and deliver advanced tools and insights, including ongoing monitoring of metrics and initiatives.
Lead highly technical, complex, cross-functional projects that drive change across a function, SBU, or corporate department.
Direct and may perform complex predictive analytics projects and cross-functional assignments in collaboration with multiple functions.
Shape the data science community and larger audiences, and lead cross-functional working groups.