Production ML Engineer - Scale & Reliability

Neura Market; San Francisco, Northern (CA, KY); Hybrid; USD 350,000 - 850,000; Full time

Job summary

Anthropic is seeking a Research Engineer for its ML Performance and Scaling team in San Francisco. You will ensure production pretrained models train reliably and efficiently at scale, bridging research and engineering across the full training stack.

You will design experiments, optimize performance, and respond to incidents during model launches while collaborating with teams in SF and London. This role combines deep technical work with operational excellence and impact on safe, beneficial AI

Qualifications

• Hands-on experience training large language models.
• Experience with JAX, TPU, PyTorch, or large-scale distributed systems.
• Willingness to be on-call during model launches and incidents.
• Strong collaboration skills across time zones and teams.

Responsibilities

• Own production pretraining pipeline components (model ops, performance, observability, reliability).
• Debug across full stack from hardware to training dynamics and eval infra.
• Design experiments to improve training efficiency and uptime.
• Respond to on-call incidents during launches and coordinate solutions.
• Build/maintain production logging, dashboards, and evaluation tools.
• Add capabilities to training codebase (long context, novel architectures).
• Collaborate with SF and London teams and with other ML groups.
• Document systems and debugging approaches for institutional knowledge.

Skills

Large language models

Distributed systems

Research engineering

On-call readiness

Cross-team collaboration

Education

Bachelor's degree in a relevant field

Tools

JAX

TPU

PyTorch

Open-source ML frameworks

Job description

Anthropic is seeking a Research Engineer for its ML Performance and Scaling team in San Francisco. You will ensure production pretrained models train reliably and efficiently at scale, bridging research and engineering across the full training stack.

You will design experiments, optimize performance, and respond to incidents during model launches while collaborating with teams in SF and London. This role combines deep technical work with operational excellence and impact on safe, beneficial AI