Research Engineer, Pretraining Scaling (AI)

Location: San Francisco, United States; this role requires working in-office 5 days per week.

Annual salary: $350,000–$850,000 USD

Company

Anthropic builds reliable, interpretable, and steerable AI systems as a public benefit corporation.

What you will do

• Own critical parts of the production pretraining pipeline, including model operations, performance optimization, observability, and reliability.
• Debug issues across hardware, networking, training dynamics, evaluation infrastructure, and the wider production stack.
• Design and run experiments to improve training efficiency, reduce step time, increase uptime, and improve model performance.
• Respond to on-call incidents during model launches and coordinate solutions across teams.
• Build logging, monitoring dashboards, evaluation infrastructure, and new training capabilities such as long-context support and novel architectures.
• Collaborate with teams in San Francisco and London, including Tokens, Architectures, and Systems, while documenting systems and debugging practices.

Requirements

• Bachelor’s degree or an equivalent combination of education, training, and experience in a relevant field.
• Hands-on experience training large language models or deep expertise with JAX, TPU, PyTorch, or large-scale distributed systems.
• Ability to combine research and engineering work, debug ambiguous multi-layer problems, and communicate clearly across teams and time zones.
• Comfort with production on-call responsibilities, extended hours during launches, and occasional evening or weekend incident response.
• Interest in responsible AI scaling and the societal impact of AI systems.
• Ability to work from the San Francisco office five days per week.

Nice to have

• Experience with JAX/TPU, PyTorch, other ML frameworks, or large-scale LLM training.
• Contributions to open-source LLM frameworks such as open_lm, llm-foundry, or mesh-transformer-jax.
• Research publications on model training, scaling laws, or ML systems.
• Experience with production ML systems, observability tools, or evaluation infrastructure.
• Background in systems engineering, quantitative work, or another field combining technical depth and operational excellence.

Culture & Benefits

• Collaborative work on a small number of large-scale AI research efforts.
• Frequent research discussions and close collaboration across disciplines.
• Competitive compensation, optional equity donation matching, generous vacation, and parental leave.
• Flexible working hours and an office designed for collaboration.
• Visa sponsorship may be available, with reasonable efforts and immigration lawyer support for successful candidates.

Hiring process

• Applications are reviewed on a rolling basis with no stated deadline.
• Candidate AI usage is governed by Anthropic’s application policy.