Model Performance Software Engineer (AI)

Location: San Francisco, CA or New York City, NY; hybrid attendance in an Anthropic office at least 25% of the time

Annual salary: $405,000–$485,000 USD

Company

Anthropic builds reliable, interpretable, and steerable AI systems through large-scale research and engineering.

What you will do

• Set technical direction for evaluation systems, research infrastructure, and internal tooling across the Claude Code team.
• Architect evaluation frameworks that measure model capabilities across diverse coding tasks and scale with the research roadmap.
• Lead infrastructure design for large-scale research experiments and make foundational architectural tradeoffs.
• Identify and deliver high-leverage engineering investments while owning critical systems from architecture through production reliability.
• Partner with research leads and product stakeholders to translate ambiguous questions into durable engineering solutions and prioritize capabilities.
• Mentor engineers, review designs, unblock peers, and raise engineering standards.

Requirements

• 10+ years of software engineering experience and a track record operating as a Staff or Principal Engineer, or equivalent.
• Experience architecting and owning complex, high-stakes pipelines, infrastructure, or platforms serving multiple teams.
• History of setting technical direction through architectural decisions, design documents, or technical strategy.
• Ability to own ambiguous problems, work across unfamiliar ML and research domains, and deliver with minimal direction.
• Strong focus on correctness, reliability, communication, and engineering standards.
• Bachelor’s degree or equivalent education, training, or professional experience in a relevant field.

Nice to have

• Experience designing or scaling evaluation frameworks for ML systems.
• Experience with reinforcement learning infrastructure, training systems, research computing, or scientific infrastructure.
• Technical leadership in high-performance environments such as trading firms, quant funds, frontier research labs, or fast-moving startups.
• Strong quantitative foundation in mathematics, physics, or a related field.
• Expertise in Python and TypeScript.

Culture & Benefits

• Collaborative research environment focused on large-scale AI safety and reliability efforts.
• Frequent research discussions and close collaboration between researchers and engineers.
• Flexible working hours and an office-based collaboration environment.
• Generous vacation and parental leave.
• Optional equity donation matching.
• Visa sponsorship is available, with immigration-lawyer support, although eligibility depends on the role and candidate.