Senior ML Platform & Infrastructure Scientist

Pinterest; Palo Alto (CA); Hybrid; USD 114,297 - 235,319; Full time

Job summary

Pinterest is seeking an experienced Data & Applied Scientist to advance ML measurement, feature understanding, and causal inference at scale. You will own end-to-end design of production ML systems and collaborate with cross-functional teams to turn research into durable platform capabilities.

The role requires 2+ years in production ML, strong Python and PyTorch skills, and experience with Airflow or similar workflow tools; in-office 3–5 times per quarter; US-based applicants only.

Qualifications

• 2+ years of hands-on experience as an applied scientist, ML engineer, research scientist or software engineer with ML production experience.
• Strong Python skills; experience with PyTorch or equivalent frameworks; familiarity with distributed compute (Spark, Ray).
• Proficiency building tools and platforms that multiply ML org impact and ensuring reproducible ML pipelines.

Responsibilities

• Translate research-grade DS workflows into production ML pipelines (proxy metrics, staleness models) using Airflow, WandB & Ray.
• Productionize causal inference methods (propensity scoring, IPW, TMLE) and build self-serve tooling for rigorous causal insights at scale.
• Partner with ML engineers and product teams to identify tooling, metrics and measurement method improvements for better model quality and business outcomes.
• Leverage metadata and engagement signals to build data-driven frameworks (feature importance, content deindexing) that improve platform efficiency.
• Design and build centralized ML platform tooling for feature/model creation, evaluation, and trust across models.

Skills

Python

PyTorch

Spark

Ray

WandB

Education

Bachelor’s/Master’s in CS

Tools

Airflow

Job description

Pinterest is seeking an experienced Data & Applied Scientist to advance ML measurement, feature understanding, and causal inference at scale. You will own end-to-end design of production ML systems and collaborate with cross-functional teams to turn research into durable platform capabilities.

The role requires 2+ years in production ML, strong Python and PyTorch skills, and experience with Airflow or similar workflow tools; in-office 3–5 times per quarter; US-based applicants only.