Data Scientist II, Infrastructure

The role

This individual contributor role sits within Pinterest's Infrastructure Data Science team and works alongside engineering and cross-functional groups. The focus is building measurement systems, metrics, and analytical frameworks that help the company understand and optimize how its platform runs across several technical areas.

What you would do

• Work with engineering teams to set, track, and raise the reliability, quality, and efficiency of internal infrastructure systems
• Create and improve metrics, dashboards, and analysis frameworks that make intricate technical systems easier to reason about
• Tighten the data base by clarifying metric definitions, checking data quality, and helping refine pipelines and measurement logic
• Run experiments and deep-dive investigations to estimate how infrastructure changes affect user experience, uptime, and business results
• Turn vague technical questions into structured analyses and concrete recommendations for engineering and platform partners
• Assist with urgent investigations and decisions about performance, reliability, cost, and measurement accuracy

What they asked for

• Master's degree in a quantitative field such as statistics or applied math, or equivalent professional experience
• Strong SQL and programming skills for analysis, including comfort working with incomplete or inconsistent data
• Hands-on experience contributing to production data pipelines, measurement systems, or data quality work
• Solid grounding in experimentation and measurement, including designing studies and interpreting results carefully
• Proven ability to structure unclear problems into analytical plans and actionable next steps
• Clear communication with engineering, product, and platform stakeholders about technical findings
• Capacity to work independently, balance longer projects against quick-turn requests, and keep momentum in a fast-changing setting
• Curiosity and a builder's approach, with motivation to improve complicated systems and build more scalable, dependable measurement foundations