Summary

Research engineer role on anthropic’s ml performance and scaling team, focused on training production pretrained frontier models reliably and efficiently at scale. You will work across the production training stack, including performance optimization, hardware debugging, experimental design, evaluation infrastructure, and on-call incident response during launches. The role is in-office 5 days per week in london.

Responsibilities

• Own critical aspects of production pretraining pipeline
• Debug issues across hardware networking and training
• Design and run experiments to improve efficiency
• Respond to on-call incidents during model launches
• Build and maintain logging monitoring and evaluation
• Add new capabilities to the training codebase
• Coordinate solutions across teams during launches