jobs

Research Engineer, Production Model Post-Training

Anthropic is hiring a Research Engineer to help train the base models that become the production Claude models. The role combines research and production engineering, with work on post-training methods that improve model capabilities, align

Company

Anthropic

Location

San Francisco, CA; New York City, NY; Seattle, WA

Status

Closed

Posted

2026-09-25T03:34:01.197+00:00

Anthropic seeks a research engineer to post-train base models that become production Claude models. The role mixes research and production work: implementing and scaling Constitutional AI and RLHF, refining post-training recipes, building fine-tuning and evaluation pipelines, measuring performance, and debugging training or model behavior. It requires strong software engineering, complex machine-learning systems experience, Python, deep-learning frameworks, and distributed computing; LLM training, fine-tuning, or evaluation experience is relevant. The posting welcomes varied experience levels, prefers senior frontier-AI engineers, mentions possible short-notice weekend incident work, and lists a $350,000–$500,000 salary.