Staff Applied Scientist, AdTech

Role overview Build applied models for insurance marketplace ranking and performance-driven media buying. The work includes choosing which carrier to show each shopper and shaping bidding and spend decisions around return on ad spend. Conversion outcomes may arrive weeks or months after a click, while sparse sessions, repeat visitors, changing markets, and the models’ influence on future data make this a challenging production problem. As an individual contributor, you will frame problems with business partners, source data, develop and validate models, and follow them through deployment and monitoring with support from an ML engineer.

Responsibilities - Build models to select likely-converting carriers for shoppers and guide paid media buying decisions. - Put quote, policy, and acquisition-cost outcomes to work in production models, accounting for delayed conversion data. - Design approaches for repeat visitors and sparse sessions while monitoring market drift and changing objectives. - Account for feedback loops in which bidding decisions influence which users and outcomes appear in future training data. - Develop marketplace bidding from early experimentation toward scaled automation and broaden the modeling scope over time. - Explain model behavior and tradeoffs clearly so business teams can rely on the resulting decisions.

Requirements - At least five years of applied data science experience with measurable business impact. - Models must have directly driven paid media spend, bidding, or ad serving; attribution or reporting by itself does not meet this requirement. - Expert Python and SQL skills, with the ability to explain model architecture, loss functions, and neural network fundamentals. - End-to-end ownership of data science work, including finding data, developing models, deployment, and monitoring. - Experience handling delayed labels, sparse features, drift, or other production challenges in decision systems.

Nice to have - Experience with live auction budgets or large-scale ranking and recommendation systems. - Knowledge of auction feedback loops, selection bias, and exploration approaches. - Practical neural network experience, including embeddings for text, creative assets, or user behavior. - AWS data science services such as ECS, S3, Athena, or SageMaker, and CI/CD deployment. - Skill in explaining technical tradeoffs to non-technical partners.

Benefits and work setup Remote-first arrangement, with most of the team working East Coast hours. The stated base salary is $175,000–$200,000 per year, paid semi-monthly; compensation also includes profit sharing and benefits. Future increases are based on company and individual performance.