Technical Operations Analyst – User Safety & Risk

About the role

OpenAI’s User Safety & Risk Operations team is looking for a Technical Operations Analyst to tackle ambiguous, high‑impact operational problems. You will work across emerging‑risk detection, workflow automation, launch readiness and quality systems, turning data and technical tools into practical solutions.

Key responsibilities

• Scope ambiguous operational problems and emerging risks, converting them into analyses, workflows, tools or decision systems.
• Analyze large, complex datasets to uncover patterns, data‑quality gaps and operational blind spots.
• Build and maintain dashboards, monitoring alerts, recurring reports and workflow automations.
• Develop systems for detecting emerging risks, including classifiers, evaluation, labeling and human‑review feedback loops.
• Support major cloud deployments by validating data pipelines, investigating anomalies and improving reporting and escalation mechanisms.
• Design quality controls such as sampling plans, golden datasets, validation holdouts and human‑review pathways.
• Leverage AI and LLM‑based tools to accelerate workflows while ensuring testing, quality assurance and human oversight.
• Document system logic, limitations and operating procedures for hand‑off to partner teams.

Required profile

• 5+ years of experience in technical operations, data analytics, trust & safety, risk operations or a related field.
• Strong judgment, discretion and resilience when handling sensitive content.
• Ability to move quickly in ambiguous environments, make thoughtful trade‑offs and create structure without perfect requirements.
• Excellent communication skills to convey complex findings to technical, operational and leadership audiences.

Required skills

• SQL for querying complex datasets.
• Python and notebook environments for data‑pipeline troubleshooting and lightweight development.
• Experience with AI/LLM tools to improve speed and scale.
• Knowledge of classifiers, evaluation, labeling programs and model‑quality measurement.
• Dashboarding and automation tooling (e.g., building alerts, monitoring systems).