Machine Learning Systems Engineer, Research Tools

Details

Anthropic is building reliable, interpretable, and steerable AI systems. The Research Tools team builds the algorithms and infrastructure researchers use to train models (production Claude and internal research models). This role focuses on improving the performance, robustness, and usability of training systems so research can progress rapidly. You will work on cutting-edge systems that train large models, implement and improve advanced techniques, and enable breakthroughs in AI capabilities and safety.

Responsibilities

• Build, maintain, and improve algorithms and systems used by finetuning researchers (RLHF and related methods).
• Improve speed, reliability, and ease-of-use of training and finetuning systems.
• Profile reinforcement learning and training pipelines to find opportunities for improvement.
• Build systems to regularly launch training jobs in test environments to detect pipeline issues quickly.
• Adapt finetuning systems to work with new model architectures.
• Create instrumentation to detect and eliminate Python GIL contention in training code.
• Diagnose and fix training slowdowns that appear after many steps.
• Implement stable, fast versions of new training algorithms proposed by researchers.
• Collaborate closely with researchers (pair programming encouraged) and support their experiments and development workflows.

Requirements

• 2+ years of software engineering experience.
• At least a Bachelor's degree in a related field or equivalent experience.
• Interest in and willingness to learn machine learning research methods.
• Results-oriented, flexible, and impact-driven mindset; comfortable picking up work beyond strict role boundaries.
• Strong collaboration and communication skills; enjoyment of pair programming and close teamwork.
• Preferred/strongly relevant experience includes:
  • High-performance, large-scale distributed systems.
  • Kubernetes.
  • Python and Python performance debugging (including GIL-related issues).
  • Implementing LLM finetuning algorithms such as RLHF.

Logistics

• Locations: San Francisco, CA and New York City, NY (United States).
• Location-based hybrid policy: staff are expected to be in one of the offices at least 25% of the time; some roles may require more on-site time.
• Visa sponsorship: Anthropic will make every reasonable effort to sponsor visas for hired candidates (not guaranteed for every role), and they retain an immigration lawyer to assist.
• Deadline to apply: None. Applications reviewed on a rolling basis.

How we're different

• Anthropic focuses on large-scale, high-impact AI research as a cohesive team. Frequent research discussions and collaboration are core to their approach. They emphasize the societal implications of AI and prioritize communication and diverse perspectives.

Benefits

• Competitive compensation (see salary range below), optional equity donation matching, generous vacation and parental leave, flexible working hours, and office space for collaboration.
• Guidance on candidate AI usage in the application process is provided.

Salary

• Annual Salary: $300,000 - $405,000 USD

Encouragement

• Candidates are encouraged to apply even if they do not meet every qualification listed. Anthropic emphasizes inclusion and representation, urging applicants from underrepresented groups not to self-select out.