Director of Data Science, Trust & Safety, Signals and Content Understanding

Pinterest Pinterest · Consumer · San Francisco, CA · Core Engineering

Director of Data Science for Trust & Safety at Pinterest, focusing on Signals and Content Understanding. This role involves setting strategy, operationalizing leading indicators, leading a data science team (including managers), partnering with Product/Engineering/Policy to ship ML capabilities, and advancing AI/ML models (including generative AI and embeddings) for content quality, user understanding, and distribution. The goal is to evolve from reactive moderation to a proactive, metrics-driven system that predicts risk and keeps Pinterest safe and inspiring. Experience with AI/agentic approaches as a force multiplier is highly valued.

What you'd actually do

  1. Set Strategy & Roadmap: Define the multi-year vision and priorities across Trust & Safety, User Understanding, and Content Understanding—establishing north-star outcomes (e.g., prevalence & reach reduction, user trust, ecosystem health, content quality, user and content understanding) and the leading indicators that drive them.
  2. Operationalize Leading Indicators: Identify, instrument, and build operating cadences around key input metrics that influence integrity and understanding—such as policy risk signals, spam/abuse rates, account authenticity, coverage and latency, precision/recall of enforcement, appeal overturn rates, content quality signals, and user feedback/friction signals.
  3. Lead a Talented Organization: Hire, develop, and inspire a world-class team of data science managers and senior ICs spanning product analytics, ML evaluation, experimentation, causal inference, and integrity measurement.
  4. Partner to Ship Impact: Work hand-in-hand with Product, Engineering and Policy teams to deliver end-to-end improvements across detection, enforcement precision, and distribution systems—translating ambiguous trust and understanding questions into shipped product/ML capabilities with measurable impact.
  5. Elevate Measurement & Decision Quality: Own measurement strategy for safety, integrity, and understanding quality, employing causal inference, experimentation, offline evaluation, and long-term value measurement to quantify trade-offs and reduce unintended consequences.

Skills

Required

  • MS or PhD in a quantitative field or equivalent experience
  • 10+ years in Data Science, Algorithmic Engineering, or ML, with significant impact in digital advertising or large-scale marketplaces
  • 5+ years of managing data science organizations, including managing managers
  • Proven ability to operate through input metrics tied to business outcomes
  • Excellent communication and influencing skills

Nice to have

  • Experience leveraging and Harnessing AI & Agentic Approaches as a Force Multiplier for Data Science, championing the adoption of AI-powered tools and agentic workflows—such as LLM-assisted analysis, automated insight generation, and AI-driven anomaly detection—to dramatically accelerate the team's ability to surface insights, iterate on measurement frameworks, and scale analytical coverage enabling teams to operate with outsized impact
  • Deep experience with causal inference, experimentation, offline evaluation, and long-term value measurement

What the JD emphasized

  • Advance AI-Driven Content & User Understanding
  • evaluate, measure, and improve AI/ML models (including generative AI and embedding-based approaches)
  • rigorous measurement frameworks
  • Experience leveraging and Harnessing AI & Agentic Approaches as a Force Multiplier for Data Science
  • championing the adoption of AI-powered tools and agentic workflows

Other signals

  • AI/ML models for content understanding
  • Generative AI and embedding-based approaches
  • Leading indicators for risk prediction
  • Metrics-driven operating system
  • Proactive risk management