Description
We're seeking a data scientist with a strong background in experimentation and statistical analysis to help us improve and iterate on our experimentation platform. The successful candidate will play a key role in improving our experiment processes at scale, leveraging their expertise to drive innovation and help make sure that Pinterest users are receiving the most thoroughly data-driven features.
With thousands of experiments running concurrently, the magnitude of our operations presents a significant opportunity for impact. If you possess a strategic mindset, proven experience in experimental design and analysis, and a passion for driving results, we invite you to join us in shaping the future of experimentation.
Responsibilities:
- Comb through the literature in experimentation to identify potential methodologies that can improve parts of our platform where we have the biggest opportunities.
- Make the process of setting up, running and evaluating experiments smoother and more repeatable for our platform users, ensuring that decisions are risk-aware and consistent.
- Write workflows to make our vast experimentation meta-data able to be leveraged by our team and outside of our team to better understand the experimentation landscape.
- Consult with product data science teams to debug, design or improve their experiments and experimentation process.
Requirements:
- PhD in a relevant field (stats, applied math, biostatistics, etc…) OR 2+ years of hands-on experience working as a data scientist or applied scientist.
- Experience working directly on experimentation problems and an awareness of state of the art methodologies.
- The ability to write clean, efficient, and scalable code that can be easily maintained and extended by other team members.
- Proficiency in software development best practices, including version control systems such as Git, to ensure efficient collaboration, code management, and reproducibility in a data science environment.
- Familiarity with workflow management tools such as Apache Airflow to create and schedule data pipelines, allowing for automated and reliable execution of machine learning workflows.