ML Systems Engineer

Position Overview

• About Anthropic: Anthropic's mission is to create reliable, interpretable, and steerable AI systems. They aim for AI to be safe and beneficial for users and society. Anthropic is a rapidly growing team of researchers, engineers, policy experts, and business leaders building beneficial AI systems.
• About the Role: This role involves building the systems that train AI models like Claude. The focus is on implementing and improving advanced machine learning techniques to create more capable, reliable, and steerable AI. The ML Systems Engineer will be responsible for the algorithms and infrastructure researchers depend on to train models, directly enabling breakthroughs in AI capabilities and safety. The primary goal is to improve the performance, robustness, and usability of these systems to accelerate research progress.

Requirements

• 4+ years of software engineering experience
• Experience with high-performance, large-scale distributed systems
• Experience with large-scale LLM training
• Proficiency in Python
• Experience implementing LLM finetuning algorithms (e.g., RLHF)
• A desire to learn more about machine learning research
• A commitment to the societal impacts of AI work

Responsibilities

• Profiling the reinforcement learning pipeline to identify opportunities for improvement.
• Building a system for regularly launching training jobs in a test environment to quickly detect problems.
• Making changes to finetuning systems to support new model architectures.
• Building instrumentation to detect and eliminate Python GIL contention in training code.
• Diagnosing and fixing issues causing training runs to slow down.
• Implementing a stable and fast version of new training algorithms proposed by researchers.
• Maintaining and improving the algorithms and systems used by finetuning researchers to train models.
• Focusing on the speed, reliability, and ease-of-use of these systems.

Compensation

• Salary: $315,000 - $425,000 USD (Annual)

Application Instructions

• Deadline to Apply: None. Applications will be reviewed on a rolling basis.

Company Information

• Anthropic's Mission: To create reliable, interpretable, and steerable AI systems.
• Team: A quickly growing group of researchers, engineers, policy experts, and business leaders.