Post-Training ML Engineer – Production Models

Anthropic Limited; San Francisco (CA); Hybrid; USD 350,000 - 500,000; Full time

Job summary

Anthropic is seeking a Research Engineer for the Post-Training team in San Francisco to train base models through the full post-training stack, ensuring alignment, safety, and production readiness. Your work blends cutting-edge research with production engineering to implement and scale post-training techniques like Constitutional AI and RLHF.

Responsibilities include designing robust pipelines, evaluating model performance, and collaborating across teams to translate new methods into

Qualifications

• Bachelor’s degree in CS or related field.
• Experience training/finetuning large language models.
• Proficiency in Python and DL frameworks.
• Experience with distributed systems and ML pipelines.

Responsibilities

• Implement and optimize post-training techniques at scale on frontier models
• Conduct research to develop and optimize post-training recipes that directly improve production model quality
• Design, build, and run robust, efficient pipelines for model fine-tuning and evaluation
• Develop tools to measure and improve model performance across various dimensions
• Collaborate with research teams to translate emerging techniques into production-ready implementations
• Debug complex issues in training pipelines and model behavior
• Help establish best practices for reliable, reproducible model post-training

Skills

Python

DL frameworks

Distributed systems

ML engineering

Education

Bachelor's degree

Tools

PyTorch

TensorFlow

Job description

Anthropic is seeking a Research Engineer for the Post-Training team in San Francisco to train base models through the full post-training stack, ensuring alignment, safety, and production readiness. Your work blends cutting-edge research with production engineering to implement and scale post-training techniques like Constitutional AI and RLHF.

Responsibilities include designing robust pipelines, evaluating model performance, and collaborating across teams to translate new methods into