Data Scientist, People Innovation

OpenAI OpenAI · AI Frontier · San Francisco, CA · Data Science

This role is for a Data Scientist within the People Innovation Labs at OpenAI, focusing on building AI-powered systems and products to improve recruiting and employee experience. The role involves defining success metrics, designing experimentation frameworks, and translating data signals into actionable decisions for scaling and intervention. The goal is to create a blueprint for how AI can enhance productivity, culture, and innovation within the organization.

What you'd actually do

  1. Define success metrics for agentic recruiting and HR systems, including leading indicators that enable weekly iteration.
  2. Design measurement and experimentation frameworks for always-on systems across every stage of the candidate and employee lifecycle — using holdouts, staged rollouts, and quasi-experimental methods when needed.
  3. Partner with PMs and engineers to instrument, evaluate, and monitor launches so every meaningful release has observability and a credible read on incremental value.
  4. Translate behavioral and model-driven signals into decisions: what to scale, where to intervene, and how to allocate human and compute attention across segments.
  5. Build repeatable decision loops (pre-launch criteria → post-launch read → next action) that convert analysis into shipped changes.

Skills

Required

  • SQL
  • Python
  • experimentation
  • causal inference
  • applied statistics
  • designing and interpreting tests in real-world, always-on environments
  • working directly with messy, incomplete behavioral data to quantify impact
  • translating results into shipped decisions
  • business judgment
  • bias toward action
  • scoping ambiguous problems
  • defining success
  • moving quickly from insight to strategy
  • communicator and partner to PMs/Engineers
  • influencing stakeholders
  • presenting recommendations to senior leadership

Nice to have

  • large language models
  • AI-assisted operations platforms
  • operational automation
  • decision systems (routing, prioritization, optimization)
  • operating in early-stage or rapidly evolving environments
  • building measurement and experimentation frameworks from scratch

What the JD emphasized

  • agentic recruiting and HR systems
  • measurement and experimentation frameworks
  • instrument evaluate and monitor launches
  • observability
  • model-driven signals
  • repeatable decision loops
  • shipped changes
  • 10+ years in a quantitative role
  • Deep grounding in experimentation, causal inference, and applied statistics
  • translating results into shipped decisions

Other signals

  • design measurement and experimentation frameworks
  • instrument evaluate and monitor launches
  • translate behavioral and model-driven signals into decisions
  • build repeatable decision loops