Model Performance Software Engineer, Claude Code

About the Role

We're looking for a Staff Software Engineer to set technical direction at the intersection of engineering and research on the Claude Code team. In this role, you'll partner directly with Anthropic's researchers and engineering leadership to shape how we measure, understand, and improve Claude's coding capabilities. You'll architect the systems, tooling, and evaluation infrastructure that determine how quickly our research can move—and you'll be accountable for the technical decisions that ripple across the team and beyond.

Responsibilities

• Set technical direction for evaluation systems, research infrastructure, and internal tooling across the Claude Code team
• Architect eval frameworks that measure model capabilities across diverse coding tasks and scale with our research roadmap
• Lead the design of infrastructure that enables researchers to run experiments at scale, and make the foundational tradeoffs that shape how the team operates for years
• Identify the highest-leverage engineering investments and drive them to completion
• Serve as a senior technical bridge between product and research, using strong product intuition to influence prioritized capabilities and metric definitions
• Mentor and raise the bar for other engineers on the team; review designs, unblock peers, and model engineering standards
• Partner with research leads to translate ambiguous research questions into durable engineering solutions
• Own critical systems end-to-end, from architecture through production reliability

Qualifications

• 10+ years of software engineering experience, with a track record of operating as a Staff or Principal engineer (or equivalent) at a high-caliber organization
• Experience architecting and owning complex, high-stakes systems (pipelines, infrastructure, or platforms orchestrating multiple components and handling significant state)
• History of setting technical direction through design docs, architectural decisions, or technical strategy
• Comfort in high-intensity environments with fast iteration cycles
• Ability to take full ownership of ambiguous, open-ended problems with minimal direction
• Power user of agentic coding tools with deep intuition about model capabilities and limitations
• Ability to dive into unfamiliar technical domains (ML systems, research workflows, novel infrastructure)

Bonus Qualifications

• Experience designing or scaling evaluation frameworks for ML systems
• Experience with reinforcement learning infrastructure or training systems
• Experience leading technical initiatives in high-performance environments (trading firms, quant funds, frontier research labs, fast-moving startups)
• Experience with research computing, scientific infrastructure, or developer platforms at scale
• Strong quantitative foundation (math, physics, or related fields)
• Expertise in Python and TypeScript