Production-Scale LLM Pretraining Engineer

anthropic; Greater London; On-site; GBP 260,000 - 630,000; Full time

Benefits offered by this job

Equity option

Flexible working hours

Vacation and parental leave

Job summary

Anthropic in London is seeking a Research Engineer for the ML Performance and Scaling team to keep production pretrained models training reliably at scale. The role sits at the boundary of research and engineering, involving the full training stack, performance optimization, and incident response.

You’ll work across SF and London teams, design experiments to improve efficiency, and contribute to deployment readiness during launches.

Qualifications

• Bachelor’s degree or equivalent in a relevant field
• Minimum years of experience as per internal level requirements
• Experience training large language models or deep expertise with JAX/TPU, PyTorch, or large-scale distributed systems

Responsibilities

• Own production pretraining pipeline components including model operations, performance optimization, observability, and reliability
• Debug and resolve complex issues across full stack from hardware to training dynamics
• Design and run experiments to improve training efficiency and model performance
• Respond to on-call incidents during model launches and coordinate solutions
• Build and maintain production logging, monitoring dashboards, and evaluation infrastructure
• Add capabilities to training codebase such as long context support or new architectures
• Collaborate with teams across SF and London, including Tokens, Architectures, and Systems
• Document systems, debugging approaches, and lessons learned

Skills

LLM training

JAX

TPU

PyTorch

distributed systems

on-call production

Education

Bachelor’s degree or equivalent

Tools

JAX

TPU

PyTorch

Observability tools

Evaluation infrastructure

Job description

Anthropic in London is seeking a Research Engineer for the ML Performance and Scaling team to keep production pretrained models training reliably at scale. The role sits at the boundary of research and engineering, involving the full training stack, performance optimization, and incident response.

You’ll work across SF and London teams, design experiments to improve efficiency, and contribute to deployment readiness during launches.