Infra Architect for Scalable AI Safeguards

Anthropic Limited; New York (NY); On-site; USD 350,000 - 500,000; Full time

Job summary

Anthropic is seeking an engineer to own and scale the Safeguards infrastructure that underpins research, experimentation, and production workloads. You will design and operate data pipelines, training and evaluation workflows, and tooling used directly by researchers to move ideas to reliable results at scale.

You will collaborate across teams to improve performance, cost, and reliability of large-scale inference while ensuring correctness as models evolve.

Qualifications

• Strong software engineering fundamentals with Python proficiency.
• Experience building and operating data-intensive or distributed systems in production.
• Experience building tooling or infrastructure used by engineers or researchers.
• Comfort working across the research-to-deployment pipeline from experiments to production.
• Ability to debug performance and correctness across an unfamiliar stack.
• Strong written and verbal communication; collaborative decisions.

Responsibilities

• Build and scale the infrastructure and data pipelines for safeguards ML research.
• Own the training, evaluation, and scoring workflows used by researchers to speed results.
• Design tooling and interfaces for researchers, including libraries and CLI tools.
• Incorporate correctness and sanity checks so results stay trustworthy as workloads evolve.
• Migrate high-value research workflows from experiments to production-grade jobs.
• Improve throughput, cost, and reliability of large-scale inference and scoring workloads.

Skills

Python

Distributed systems

Infra tooling

Research-to-deployment

Debug performance

Communication

Education

Bachelor’s degree in a relevant field

Tools

ML infra tooling

Data pipelines tooling

Job description

Anthropic is seeking an engineer to own and scale the Safeguards infrastructure that underpins research, experimentation, and production workloads. You will design and operate data pipelines, training and evaluation workflows, and tooling used directly by researchers to move ideas to reliable results at scale.

You will collaborate across teams to improve performance, cost, and reliability of large-scale inference while ensuring correctness as models evolve.