Lead Data Scientist: Claims and Service DS Knowledge & Development Solutions

Description The Claims and Service AI Enablement team is hiring a Data Scientist to own Generative AI and machine learning support for Knowledge & Development Solutions (KDS) .

This role will partner with KDS stakeholders to design, build, evaluate, and scale AI/ML solutions across knowledge management, training, content delivery, and employee enablement.

KDS helps Liberty Mutual deliver the right content to the right audience at the right time through contextual knowledge, guided learning, learning experience design, training operations, and knowledge-centered support.

This role will apply predictive modeling, Rapid experimentation & prototyping, GenAI, RAG, traditional ML, and evaluation to improve how employees find, learn, and use information.

**Candidates who live within 50 miles of Boston, MA; Portsmouth, NH; Seattle, WA; Columbus, OH; or Plano, TX will follow a hybrid schedule, coming into the office two days per week. Otherwise, this role is remote with occasional travel.

** Responsibilities: Own the Generative AI and ML portfolio for KDS in partnership with KDS leaders Identify, prioritize, and deliver AI/ML use cases that improve knowledge discovery, content effectiveness, learning outcomes, and employee enablement.

Design, build, and evaluate predictive models, GenAI applications, semantic search, RAG solutions, and analytical tools. Apply full-stack data science practices from problem framing and data exploration through modeling, evaluation, deployment partnership, measurement, and optimization.

Partner with KDS, Technology, Engineering, Architecture, UX, Compliance, and business stakeholders to move solutions from concept to production-minded implementation.

Define and implement measurement strategies to assess model performance, adoption, operational efficiency, content quality, training impact, and business outcomes. Provide technical mentorship and thought leadership on AI/ML solution design, evaluation, experimentation, and scalable implementation p