jobs

Research Engineer, Pretraining Scaling

About Anthropic Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers,

In inglese

Company

Anthropic

Location

San Francisco, CA

Status

Open

Posted

2025-09-30T20:35:48+00:00

Anthropic is hiring a research engineer for its ML Performance and Scaling team, which trains the company's production pretrained models. The person owns parts of that pretraining pipeline: performance optimization, observability, reliability, hardware and networking debugging, experiments that cut step time and raise uptime, plus logging, monitoring and evaluation infrastructure. On-call incident response during launches is expected, with extended hours. Anthropic asks for hands-on large language model training experience or deep JAX, TPU, PyTorch or distributed-systems expertise. Work is in-office five days a week in San Francisco, alongside teams there and in London.