Machine Learning Infrastructure Engineer, Safeguards Research

About the role

Anthropic’s Safeguards team builds systems that detect and mitigate misuse of AI models. We need an engineer to own the infrastructure that powers this research, enabling fast iteration and reliable production‑grade detections. You will bridge the gap between experimental research and scalable deployment.

Key responsibilities

• Design, build, and scale data pipelines and infrastructure for Safeguards machine‑learning research.
• Own training, evaluation, and scoring workflows to shorten the time from idea to result.
• Create libraries, command‑line tools, and interfaces that researchers can use without deep systems knowledge.
• Implement correctness and sanity checks to keep results trustworthy as models evolve.
• Transition high‑value research workflows into reliable production jobs.
• Improve throughput, cost efficiency, and reliability of large‑scale inference and scoring workloads.
• Collaborate closely with researchers and engineers to anticipate workflow changes and design proactive solutions.

Required profile

• Strong software engineering fundamentals with hands‑on coding ability.
• Experience building and operating data‑intensive or distributed systems in production.
• Proven track record of delivering tooling or infrastructure used by other engineers or researchers.
• Comfort working across the full research‑to‑deployment pipeline, from exploratory experiments to production systems.
• Ability to debug performance and correctness problems in complex workloads.

Required skills

• Python