Staff Software Engineer, Developer Productivity (Dev Environments) - Claude Code
AnthropicAbout Anthropic
Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.
About the role
Every minute an Anthropic engineer or researcher spends waiting on a container to boot, a dependency to resolve, or CI to surface a failure that could have been caught locally is a minute not spent on frontier AI safety. The Developer Productivity team owns the inner development loop for the people building Claude — and increasingly, for Claude itself as an engineering collaborator.
We're looking for a Staff Software Engineer to own our development environment end to end. You'll make spinning up a fresh environment fast, design the isolation boundary that keeps local experimentation safely separated from shared research and production infrastructure, and build the pre-push validation layer that catches problems before they ever reach CI. You'll also help shape how our repository and platform topology evolves as the company scales, partnering closely with the infrastructure teams who own the substrate you build on.
This team works the way it expects the rest of Anthropic to work: with Claude in the loop. Agentic coding is both how we operate day to day and a core part of what these environments need to support, so we're looking for someone who already builds this way and has informed opinions about where the leverage actually is.
Key responsibilities
-
Own the local and hosted development environment end to end — container lifecycle, dependency provisioning, hot reload, and the single command an engineer runs to start working
-
Drive down cold-start time for fresh development environments and keep it low as the codebase grows
-
Design and implement the environment isolation model (sandboxes, ephemeral environments, namespace separation) that lets engineers experiment freely without risk to shared systems
-
Build and maintain the pre-push validation surface so failures are caught on the engineer's machine, not in CI
-
Partner with platform, delivery infrastructure, and tooling teams to shape the repository and service topology that best supports a fast inner loop
-
Act as a technical lead across team boundaries — gathering require