Staff+ Software Engineer, Safeguards ML Infrastructure

Overview

Software Engineer for the Safeguards ML Infrastructure team, designing and running production infrastructure for Claude’s safety systems.

What you'll do

• Design, build, and deploy backend services critical to token sampling and generation.
• Own and operate production serving infrastructure across multiple deployment platforms (1P, AWS Bedrock, GCP Vertex).
• Define and maintain SLOs, and build observability and alerting for early problem detection.
• Lead incident response and participate in on-call/operational-duty rotations.
• Reduce operational toil by building automation, tooling, and self-serve workflows.
• Build and maintain a safety registry with full provenance for production deployments.
• Implement automated post-deploy validation and work with ML researchers to productionize safety techniques.

What you'll need

• Proficiency in Python.
• Experience deploying and operating on cloud platforms (AWS, GCP).
• Designed, built, and operated high QPS systems at global scale.
• Strong distributed systems foundation, including SLO management under load.
• Meaningful on-call experience for production incidents and postmortem-driven improvements.
• Hands-on experience deploying at scale on cloud platforms (AWS, GCP).

Details

• Annual compensation range: $320,000–$485,000 USD.
• Work mode: location-based hybrid policy; staff expected in an office at least 25% of the time.
• Visa sponsorship: sponsored for some roles; not guaranteed for every candidate/role.