Engineering Manager, Model Flywheel

OpenAI OpenAI · AI Frontier · San Francisco, CA · Applied AI

Engineering Manager for the ChatGPT Model Capabilities and Deployment team, focusing on transforming model advancements into user experiences through reliable serving, rapid experimentation, safe deployment, and continuous improvement. Responsibilities include managing model experimentation, deployment, and measurement systems, including harness, context management, system prompts, multi-tier model experiences, experiment self-serve capabilities, automated guardrails, rollout automation, capacity management, health monitoring, and end-to-end measurement systems (evals, grader signals, user feedback). Requires experience leading engineering teams, shipping production AI systems at scale, and understanding model-driven product development and deployment lifecycles.

What you'd actually do

  1. Elevate and consolidate ChatGPT’s harness, context management, and system prompt frameworks.
  2. Drive expansion and improvement of multi-tier model experiences.
  3. Support and scale self-serve experiment capabilities and automated guardrails.
  4. Lead model rollout automation, capacity management, and health monitoring.
  5. Shape end-to-end measurement systems (evals, grader signals, user feedback, etc.).

Skills

Required

  • Proven experience leading engineering teams in complex, cross-functional environments.
  • Demonstrated success shipping production systems at scale (ideally for AI or large backend services).
  • Deep understanding of model-driven product development, deployment lifecycle, and measurement tooling.
  • Excellent communication and collaboration skills—experience interfacing directly with engineering, research, and product stakeholders.

Nice to have

  • Prior involvement with large language models, distributed infrastructure, or experimentation platforms is a plus.

What the JD emphasized

  • shipping production systems at scale
  • AI or large backend services
  • model-driven product development
  • deployment lifecycle
  • measurement tooling
  • large language models
  • distributed infrastructure
  • experimentation platforms
  • AI reliability
  • safety
  • user impact
  • global scale

Other signals

  • ChatGPT user experiences
  • reliable serving
  • rapid experimentation
  • safe deployment
  • continuous improvement
  • model advancements
  • model capabilities
  • measurement systems
  • user signals
  • launch scorecards
  • feedback loops
  • experiment automation
  • automated guardrails
  • capacity management
  • health monitoring
  • end-to-end measurement
  • production systems at scale
  • AI or large backend services
  • model-driven product development
  • deployment lifecycle
  • large language models
  • distributed infrastructure
  • experimentation platforms
  • AI reliability
  • safety
  • user impact
  • global scale