Drimva

Anthropic

Staff+ Software Engineer, Claude App Infrastructure

$320K–$485K

San Francisco, CA +3 · Staff · Full-time · Posted Jan 13, 2026

Location
San Francisco, CA
New York City, NY
Seattle, WA
San Francisco, California, United States
Experience
Years required correlate with internal job level requirements; no specific number stated.
Education
Bachelor’s degree or an equivalent combination of education, training, and/or experience; relevant field demonstrated through coursework, training, or professional experience.
Visa sponsorship
Visa sponsorship is offered, but may not be possible for every role or candidate; the company says it will make reasonable efforts to obtain a visa for candidates it offers.

About the job

Build infrastructure for claude.ai’s agentic capabilities, enabling Claude to execute tasks and use tools securely, reliably, and at scale.

What you'll do

  • Build sandboxed compute environments and state management for long-running agent tasks.
  • Develop secure delegated authentication and authorization, plus observability and debugging tools for agent execution.
  • Partner with product and research teams to define and ship capabilities.

What they ask for

  • Experience building distributed systems, infrastructure, or platform services at scale.
  • Comfort building cloud-native infrastructure on GCP, AWS, or Azure.
  • Experience with containers, sandboxing, or secure execution environments.
  • Write maintainable code in Python, Go, Rust, or a similar language; care about security and safe failure.

Nice to have

  • Experience with multi-tenant execution platforms, serverless infrastructure, security engineering, sandboxing, or isolation technologies.
  • Familiarity with workflow orchestration, durable execution patterns, or low-level systems.

Pay

$320,000 — $485,000 USD annual base salary

Tools

GCP, AWS, Azure, gVisor, Firecracker, V8 isolates, Python, Go

AI focus

Build agentic infrastructure that lets Claude execute code, use tools, access external services, and act on users’ behalf.

Model-written summary · the original posting is authoritative

First seen Oct 8, 2026 · Last verified Nov 5, 2026 (UTC)