Data Scientist III

Modeling & Technical Execution:

Own the end-to-end data science lifecycle for moderately complex models and significant project components — spanning data ingestion, feature engineering, modeling, validation, deployment, monitoring, and retraining.

Apply expertise across several core areas of machine learning and statistics (e.g., gradient-boosted models, deep neural networks, time series, causal inference concepts, experimentation design), selecting appropriate methods for complex data science problems.

Write efficient, modular, well-tested code for data processing, feature engineering, and model training/inference, leveraging distributed tooling (e.g., Vertex AI pipelines, Dataflow, BigQuery) where appropriate.

Design and implement robust validation frameworks for complex experiments and models, accounting for potential biases and real-world performance.

Troubleshoot complex model performance issues, data anomalies, and code bugs effectively with little guidance.

Execution & Collaboration:

Define analytical approaches and scope data science projects for moderately complex or ambiguous business problems.

Partner with product managers and stakeholders to define success metrics and experiment goals, and to translate marketplace problems into data science solutions.

Lead the design and analysis of experiments (e.g., A/B tests, switchback) for your projects, and interpret complex model results and experimental outcomes with a focus on actionable insights and business outcomes.

Proactively identify opportunities within your domain where data science can provide significant value, and initiate exploration.

Follow and help improve established team processes for coding standards, documentation, reproducibility, and experimentation.

Mentorship & Influence:

Mentor DS I and DS II scientists, providing technical guidance, reviewing code, analyses, and models, and supporting their growth in analytical and modeling skills.

Influence technical decisions within the team regarding modeling choices, validation strategies, and tooling through well-reasoned arguments and expertise.

Drive improvements to team standards, data science best practices, and analytical rigor; take ownership of specific team practices or technical components (e.g., a feature store component, leading experimentation reviews).

Educate stakeholders on the capabilities and limitations of data science models, and clearly explain complex methodologies and findings to both technical and non-technical audiences.

Participate actively in recruiting, providing high-quality, graded interview feedback for candidates up to this level.