Role overview
As a core member of the Data Science team, the Principal Data Scientist will lead the design and development of high-impact machine learning systems that power CarGurus’ core products and strategic initiatives. You will own models and data pipelines end-to-end from problem framing and experimentation through production deployment, monitoring, and continuous improvements. Potential areas of support and ownership include Recommendations, Search Ranking, Instant Market Value algorithms, and new ML-driven product capabilities.
What you'll do
- Own the end‑to‑end design and implementation of production ML models and systems, including model architecture, feature strategy, and evaluation methodology.
- In partnership with Data and Analytics teams, develop and maintain data pipelines to supply training and inference data for models, using SQL and Snowflake.
- Collaborate with engineering leaders on system and API design to ensure ML solutions meet requirements for latency, reliability, observability, and maintainability in production.
- Apply best practices for experimentation and model evaluation, including offline metrics, A/B testing design, and post‑launch analysis; coach other data scientists in applying these practices rigorously.
- Communicate solutions to stakeholders through written documentation, demos and presentations, and data visualizations tailored to both technical and non‑technical audiences.
What you'll bring
- 7+ years of experience in Data Science or Machine Learning roles, with a consistent track record of shipping, owning, and iteratively improving production ML systems that drive material business impact.
- Deep expertise in Machine Learning techniques for supervised and unsupervised learning across structured and unstructured datasets. Comprehensive knowledge of, and real-world experience with, measurement, evaluation, and testing of models.
- Proven experience deploying and maintaining machine learning services in production, ideally in a cloud environment (e.g. AWS, SageMaker, Snowflake).
- High proficiency in Python (or a similar language widely used in the data science community) and SQL, plus experience establishing coding, testing, and reproducibility standards (e.g., shared libraries, experiment tracking, templates).
- Ability to communicate technical details, trade‑offs, and analytical findings to audiences ranging from engineers to senior business leaders, using clear narratives and data‑driven recommendations.
- Advanced degree (or proven experience) in Computer Science, Data Science, Mathematics, or any quantitative science which makes use of advanced data analytics or statistical or machine learning techniques.