Data Scientist, Support

OpenAI OpenAI · AI Frontier · San Francisco, CA · User Operations

Data Scientist focused on analyzing user support data to improve customer experience and product adoption. This role involves building dashboards, establishing metrics, leveraging LLMs for classification, and collaborating with Data Engineering and Data Science teams. The goal is to provide actionable insights and scalable reporting for internal stakeholders and BPO partners.

What you'd actually do

  1. Explore large support and product datasets to uncover trends, volume drivers, and user‑experience pain points, distilling findings into clear, actionable narratives.
  2. Build, enhance, and maintain self‑serve dashboards and reporting tools, enabling non‑technical teams to answer their own data questions.
  3. Establish a unified metrics taxonomy for service‑health and performance—and build automated data‑sharing pipelines and scorecards with our BPO partners to ensure everyone operates from the same real‑time view of success
  4. Leverage LLMs to build bespoke classifiers that automatically label and segment inbound volumes—powering smarter routing, richer self‑serve insights, and swifter root‑cause analysis.
  5. Partner with Data Engineering to ensure reliable pipelines, implement data‑quality checks, and document sources of truth.

Skills

Required

  • SQL
  • Python
  • R
  • BI dashboard design and maintenance
  • LLM prompt engineering
  • LLM fine-tuning
  • Data analysis
  • Data visualization
  • Cross-functional communication

Nice to have

  • Customer support experience
  • Operations experience
  • Familiarity with support metrics (SLAs, FCR, deflection)
  • Experience with LLM classifiers
  • Experience with Retool

What the JD emphasized

  • 8 + years in analytics, business intelligence, or data science
  • Expert-level SQL skills
  • proficiency in Python or R
  • Hands-on experience designing and maintaining BI dashboards
  • Hands-on experience fine-tuning or prompt‑engineering LLMs
  • strong bias for automation
  • bias toward action over theoretical perfection

Other signals

  • customer support data
  • self-serve analytics
  • LLM classifiers
  • real-time insights