Staff Data Scientist, Lead Generation

Role overview Build and operate machine-learning models that guide how auto insurance shoppers are matched with carriers and how paid media is allocated. This hands-on individual-contributor role spans problem framing, data discovery, modeling, validation, deployment, and monitoring, with an ML engineer supporting production work. A central challenge is learning from delayed outcomes, sparse sessions, repeat visitors, shifting markets, and bidding decisions that influence future training data.

Responsibilities - Develop models that select the carrier most likely to convert each shopper and support performance-based media buying. - Bring carrier outcomes, including quotes, policies, and acquisition costs, into production models. - Own the full modeling lifecycle: define business problems, source data, build and validate models, and monitor their results after deployment. - Address delayed conversion labels, sparse user activity, repeat shoppers, market drift, and feedback effects from auctions and bids. - Work with an ML engineer on production implementation and broaden modeling work beyond auto insurance over time. - Establish approaches for user matching, automated bidding, and reliable model performance as objectives change.

Requirements - Five or more years of applied data science experience with measurable business impact. - Experience building models that directly informed paid media spending, bidding, or ad serving. - Strong understanding of model design, including loss functions, architecture choices, and neural network fundamentals; able to explain and substantiate work personally built. - Experience with late-arriving labels, sparse features, drift, or the ways decisions can shape subsequent training data. - Ability to own models end to end, from data discovery through deployment and monitoring. - Expert Python and SQL skills.

Nice to have - Experience buying media in live auctions, or building large-scale ranking and recommendation systems. - Knowledge of auction feedback loops, selection bias, and exploration; practical neural-network work such as embeddings. - Ability to explain technical tradeoffs to nontechnical partners; familiarity with AWS data science services such as ECS, S3, Athena, or SageMaker, and CI/CD.

Benefits and work setup The team is remote-first, and most colleagues work East Coast hours. The source describes an outcome-focused environment with direct, respectful feedback and a fast-changing pace.