Machine Learning Infrastructure Engineer (AI)

TL;DR

Machine Learning Infrastructure Engineer (AI): Building and scaling the infrastructure and data pipelines for Safeguards research to detect and mitigate AI misuse with an accent on reducing iteration time for researchers. Focus on designing high-performance tooling for large-scale inference, scoring workloads, and bridging the gap between research experiments and production-grade systems.

Location: Hybrid (San Francisco, CA or New York City, NY). Staff are expected to be in the office at least 25% of the time.

Salary: $350,000 - $500,000 USD

Company

Anthropic is a public benefit corporation focused on creating reliable, interpretable, and steerable AI systems to ensure they are safe and beneficial for society.

What you will do

• Build and scale the infrastructure and data pipelines supporting Safeguards machine learning research.
• Own training, evaluation, and scoring workflows to minimize the time between research ideas and results.
• Design libraries and command-line tools that abstract complex system internals for researchers.
• Implement correctness and sanity checking within the stack to maintain result trustworthiness as models evolve.
• Transition high-value research experiments into reliable, production-grade jobs.
• Optimize the throughput, cost, and reliability of large-scale inference and scoring workloads.

Requirements

• Proficiency in Python and strong software engineering fundamentals.
• Experience building and operating production-grade data-intensive or distributed systems.
• Track record of developing infrastructure or tooling used as a dependency by other engineers/researchers.
• Ability to work across the entire research-to-deployment pipeline.
• Experience debugging performance and correctness issues across unfamiliar technical stacks.
• Must be based in or able to work from San Francisco, CA or New York City, NY.

Nice to have

• Experience with high-performance, large-scale ML systems.
• Familiarity with transformers, language modeling, and model internals.
• Knowledge of GPU/accelerator programming or inference optimization.
• Experience building experiment tracking, caching layers, or evaluation harnesses.
• Interest in AI misuse risks and mitigations.

Culture & Benefits

• Competitive compensation with optional equity donation matching.
• Generous vacation and parental leave policies.
• Flexible working hours and collaborative office environments.
• Visa sponsorship availability for qualified candidates.
• Commitment to diverse perspectives and inclusive hiring practices.