Technical Program Manager – Safeguards Infrastructure & Evaluations

About the role

Anthropic is building reliable, interpretable and steerable AI systems. As a Technical Program Manager for Safeguards Infrastructure and Evaluations, you will own the operational health of the safety‑critical pipelines that keep AI models safe in production, ensuring incidents are handled quickly and long‑term reliability improves.

Key responsibilities

• Lead the Safeguards Engineering ops review cadence, surfacing incidents, reliability trends and coordinating decision‑making.
• Own incident tracking across teams, drive post‑mortem creation and ensure all action items are completed.
• Define, negotiate and monitor Service‑Level Objectives (SLOs) with partner teams such as Inference and Cloud Inference.
• Coordinate major platform investments, including migrations and evaluation‑platform enhancements.
• Collaborate with engineers to triage issues, assess safety‑critical impact and prioritize remediation.

Required profile

• Strong program‑management experience in a technical environment.
• Deep understanding of complex infrastructure systems and ability to diagnose issues quickly.
• Proven track record of managing incident response and post‑mortem processes.
• Excellent communication skills for cross‑team coordination and stakeholder alignment.
• Passion for AI safety and reliability.

Required skills

• Incident response and management.
• Post‑mortem execution and follow‑through.
• Service‑Level Objective (SLO) definition and monitoring.