Pinterest is the world's leading visual search and discovery platform, serving over 500 million monthly active users globally on their journey from inspiration to action. As we scale experiences in a complicated ecosystem, ensuring they are safe, fair, and trustworthy is paramount. We are looking for a Senior Data Scientist to help lead Pinterest's Trust and Safety mandate by designing the foundations for measuring the prevalence of unsafe content across the platform. In this role, you will design and build sampling frameworks, complex data aggregations, and measurement methodologies to track Trust & Safety policy violations across complex, multi-component user interactions. You will work in a highly collaborative and cross-functional environment, partnering with ML Engineers, Trust & Safety Ops, subject matter expert teams, and Product Managers. The results of your work will directly influence platform safety metrics, policy compliance, and executive-level visibility into platform health. What you'll do: - Design and develop ML-assisted sampling techniques, applying expertise in statistical methods to accurately measure the prevalence of unsafe content, treating complex multi-component interactions as distinct measurement units. - Apply rigorous statistical methods, drawing on knowledge of all kinds of sampling methods and their proper statistical application for complicated use cases, to calculate prevalence rates for specific Trust & Safety policy violations (e.g., Adult content, Self-harm, Harassment, Misinformation) and to further expand and improve prevalence measurement. - Build large-scale data pipelines to aggregate Pinner-generated queries, system responses, and recommended Pin images into a unified format for human and ML-based safety labeling. - Partner cross-functionally to orchestrate "Offline" dashboards and robust "Online" production workflows for continuous safety monitoring. - Collaborate closely with Trust & Safety teams to translate written safety policies into unified LLM prompts, coordinate BPO labeling queues, and calibrate labeler decision quality. - Define and evangelize what constitutes high-quality content across Pinterest's surfaces, building rigorous, scalable statistical frameworks to measure and continuously monitor content quality at a platform level. - Analyze and model the end-to-end content distribution funnel to inform how high-quality content is selected, ranked, and surfaced to hundreds of millions of Pinners, creators, advertisers, and merchants. - Lead the development of improved methodologies for evaluating the quality of links to external websites surfaced on Pinterest, partnering with Engineering and Policy teams to operationalize findings. - Develop best practices for instrumentation and experimentation, and instrument methodology to improve the sensitivity of existing metrics, across both the Trust & Safety and Content Quality domains. - Design reusable tooling and workflows for ongoing metrics monitoring and reporting across both mandates. - Leverage AI to seek faster execution (i.e. draft, prototype, outline) and explore alternative options (i.e. iterate, compare approaches) - Leverage AI to synthesize information (summarize, distill themes) and automate repeatable tasks (documentation, reporting, QA checks) What we're looking for: - 5+ years of experience analyzing data in a fast-paced, data-driven environment with proven ability to apply scientific methods to solve real-world problems on web-scale data. - Extensive experience solving analytical problems using quantitative approaches in Machine Learning, Statistical Modeling, Forecasting, Econometrics, or related fields, with a proven record of researching and implementing advanced methods on real-world measurement problems. - Strong interest and hands-on experience in platform safety, prevalence measurement, content quality measurement, adversarial testing, responsible data measurement, or Trust & Safety. - Deep familiarity with the measurement challenges of a complex ecosystem, including statistical interpretation of data across multimodal and unstructured content types. - Experience designing and calibrating measurement frameworks, managing complex logging tables (e.g., user/interaction/component data), and defining directional success metrics. - Experience using machine learning and deep learning frameworks, such as PyTorch, TensorFlow, or scikit-learn. - Strong quantitative programming (Python) and data manipulation skills (SQL/Spark); experience with complex ML pipelines and up-sampling. - A scientifically rigorous approach to analysis and data, with a well-tuned sense of skepticism, high intellectual curiosity, and attention to detail. - Ability to drive ambiguous measurement projects end-to-end, overcoming unstructured policy dependencies with high ownership. - Excellent written and verbal communication skills, with the ability to advocate for decision quality before releasing metrics to executive leadership, and to explain learnings to both technical and non-technical partners. - A team player who is able to partner with cross-functional leadership-Product, Engineering, Design, Research, Trust & Safety Ops, and Data Engineering-to quickly turn insights into action. - Demonstrated ability to use AI to improve speed and quality in your day-to-day workflow for relevant outputs. - Strong track record of critical evaluation and verification of AI-assisted work (e.g., testing, source-checking, data validation, peer review). - High integrity and ownership: you protect sensitive data, avoid over-reliance on AI, and remain accountable for final decisions and deliverables. - Bachelor's, Master's degree, or PhD in a relevant field such as Data Science, or equivalent experience This job posting is for an open vacancy. Please note that the company utilizes artificial intelligence to screen applicants for the positions. Relocation Statement: - This position is not eligible for relocation assistance. Visit our PinFlex page to learn more about our working model. In-Office Requirement Statement: - We recognize that the ideal environment for work is situational and may differ across departments. What this looks like day-to-day can vary based on the needs of each organization or role. - This role will need to be in the office for in-person collaboration 1-2 times every 6-months and therefore can be situated anywhere in Ontario. #LI-HYBRID #LI-CH1