AI Safety Measurement Program Lead

OpenAI, Inc.; San Francisco (CA); Presencial; USD 180 000 - 240 000; Tempo integral

Vantagens oferecidas por esta oferta de emprego

Relocation assistance

Resumo da oferta

OpenAI, Inc. is seeking a Safety Measurement Research Program Manager to drive the strategy for safety measurement and evaluations across research and production.

You will work with researchers, data scientists, policy owners, and product teams to shape data, dashboards, tools, and workflows, prioritizing problems and delivering across problem definition to rollout. The role combines measurement expertise, product judgment, and cross-functional execution, with relocation assistance available for

Qualificações

• 6+ years in research program management, technical program management, product management, or related roles.
• Strong foundation in data science, statistics, economics, or related field.
• Experience with measurement or evaluations in AI safety or trust & safety domains.

Responsabilidades

• Define strategic priorities and roadmaps to improve safety measurement and evaluation.
• Partner with research, data science, engineering, policy, and product teams to translate vision into programs.
• Drive integration of AI safety research into safety measurement platforms; navigate tradeoffs and blockers.
• Develop clear success criteria for safety measurement and evals.
• Represent quantitative progress on safety to senior leadership.

Conhecimentos

AI safety

Data science foundations

Measurement & evaluations

Cross-functional leadership

Strategic roadmaps

Descrição da oferta de emprego

OpenAI, Inc. is seeking a Safety Measurement Research Program Manager to drive the strategy for safety measurement and evaluations across research and production.

You will work with researchers, data scientists, policy owners, and product teams to shape data, dashboards, tools, and workflows, prioritizing problems and delivering across problem definition to rollout. The role combines measurement expertise, product judgment, and cross-functional execution, with relocation assistance available for