Research Engineer, Production Model Post-Training

About the role

Anthropic is looking for a Research Engineer to work on the post‑training stack that turns base models into the production Claude models used by customers. The role sits at the intersection of cutting‑edge AI research and large‑scale engineering, focusing on techniques such as Constitutional AI and RLHF to improve model alignment, safety and performance.

Key responsibilities

• Implement and scale post‑training methods on frontier language models.
• Design, build and maintain robust pipelines for fine‑tuning, evaluation and deployment.
• Research and develop new post‑training recipes that directly boost production model quality.
• Create tools to measure model performance across multiple dimensions.
• Collaborate with research teams to translate emerging techniques into production‑ready code.
• Debug complex issues in training pipelines and model behavior.
• Establish best practices for reliable and reproducible model post‑training.

Required profile

• Comfortable working in a fast‑moving, ambiguous environment with multiple urgent priorities.
• Strong problem‑solving skills for time‑sensitive debugging.
• Experience building large‑scale ML systems and working with distributed computing.
• Background in training, fine‑tuning or evaluating large language models.
• Interest in AI safety and responsible deployment.

Required skills

• Python programming.
• Deep learning frameworks (e.g., PyTorch, TensorFlow).
• Distributed computing and high‑performance computing.
• Large‑scale model training and fine‑tuning.