ML Systems Engineer: RL Training & AI Safety

Anthropic; Seattle (WA); Hybrid; USD 520,000 - 850,000; Full time

Benefits offered by this job

Equity donation matching

Generous vacation

Parental leave

Flexible working hours

Office space

Job summary

Anthropic is seeking an ML Systems Engineer on the Reinforcement Learning Engineering team to build the algorithms and infrastructure used to train Claude-like models. You will focus on performance, robustness, and usability to accelerate research progress and model capabilities.

You will work on production RLHF training systems, improve tooling, and collaborate with researchers. This role blends research support with engineering ownership, in a hybrid office setting.

Qualifications

• 4+ years of software engineering experience.
• Like working on systems and tools that make other people more productive.
• Are results-oriented, with a bias towards flexibility and impact.
• Pair programming is valued and practiced.
• Interest in machine learning research.
• Care about the societal impacts of your work.

Responsibilities

• Build, maintain, and improve algorithms and systems that researchers use to train models.
• Improve speed, reliability, and ease-of-use of these systems.
• Support and empower our research team to progress quickly and safely.
• Profiling and optimizing RL training pipelines and infrastructure.

Skills

4+ years software engineering

Pair programming

Results oriented

Interest in ML research

Societal impact

Productivity tooling

Education

Bachelor’s degree or equivalent

Tools

Python

Job description

Anthropic is seeking an ML Systems Engineer on the Reinforcement Learning Engineering team to build the algorithms and infrastructure used to train Claude-like models. You will focus on performance, robustness, and usability to accelerate research progress and model capabilities.

You will work on production RLHF training systems, improve tooling, and collaborate with researchers. This role blends research support with engineering ownership, in a hybrid office setting.