Summary

Staff+ software engineer on anthropic’s safeguards team, building foundational infrastructure for safety, oversight, and intervention mechanisms. You will develop systems to monitor models, detect unwanted behaviors, and prevent disallowed use, including data storage/management, evaluation metrics, and tooling for human and agentic review. The role follows a location-based hybrid policy with at least 25% office time.

Responsibilities

• Develop foundational safeguards infrastructure and tooling
• Build data storage and management systems
• Create metric and evaluation systems
• Run safeguards systems day-to-day
• Build multi-layer defenses for real-time safety
• Reduce human intervention and oversight needs