Staff+ Software Engineer, Platform Acceleration

Anthropic•San Francisco, NY

1d•Hybrid

About The Position

Platform Acceleration builds the internal tools that close that gap for Anthropic's product engineers, who own the Claude API, our SDKs, and the rest of the Claude Platform. We work like a small product team whose customers are other Anthropic engineers. We're looking for Staff+ engineers with a backend and infrastructure background who care about developer experience and have built tools other engineers rely on.

Requirements

Experience building and operating large-scale backend services or distributed systems in production
Proficiency in Go, Python, Rust, or TypeScript
Deep experience with developer tooling, CI/CD, test infrastructure, or deployment systems
Experience with cloud infrastructure and container orchestration
Experience leading large, complex technical projects across teams, from an ambiguous problem to a tool people use
Clear written and verbal communication, with experience supporting internal partner teams
Interest in Anthropic's mission of building safe, beneficial AI

Nice To Haves

Experience at a startup or on a similarly small team, taking a product from zero to one and talking directly to your users
Substantial experience with agents that do engineering or other technical work
Experience building widely adopted CLI tools, developer-facing services, GitHub automation, or open source developer tools
Depth in progressive delivery, observability, VM orchestration, sandboxed execution, or workflow orchestration
Experience working in a large monorepo
Enough full-stack ability to put a usable interface on your tools
A point of view on how software engineering changes as agents write more of the code

Responsibilities

Own the technical strategy and roadmap for your area, and make sure its hardest problems get solved, by you or through others
Stay hands-on: design, build, and operate the tools that help Platform teams ship code faster
Define what it takes to trust a change, including changes written by Claude, and build the tooling that holds that bar before and after merge
Build agents that take on repetitive engineering work, such as diagnosing failures, fixing broken tests, and running migrations
Embed with partner teams to find their biggest sources of friction, validate a solution with them, and turn the fix into a product other teams can adopt
Measure developer experience and adoption, and use the data to decide what to build next
Move between building new tools and improving the shared systems teams already depend on, as priorities shift