Software Engineer, Claude Design

About 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.

About the role
Claude Design is the newest product from Anthropic Labs, the internal accelerator behind Claude Code and MCP. It lets anyone collaborate with Claude to create polished visual work: interactive prototypes, product mockups, slides, one-pagers, and marketing collateral. Start from a prompt, an image, or your own codebase, and Claude builds a first version that you refine through conversation, inline comments, direct edits, and live adjustment controls, then hand off to Claude Code to build for real. We launched in research preview in April 2026, and the product is new enough that the engineers who join now will define what it becomes.

We're looking for Software Engineers to help build and shape Claude Design. This is a craft-heavy, frontend-leaning role: you'll create the canvas, real-time editing, design-system inference, and AI-driven generation experiences that make AI-generated design feel like a tool people reach for first, not a demo. You'll work at the intersection of cutting-edge research and real-world application, rapidly building and testing new experiences, partnering directly with researchers and users.

Responsibilities:

• Build and ship Claude Design's core surfaces (the design canvas, real-time editing, inline comments, and the live adjustment controls Claude generates), shipping early and often.
• Collaborate closely with research teams to turn frontier vision and generation capabilities into intuitive, high-craft design experiences.
• Work directly with designers, internal test users, and external partners to gather feedback, iterate quickly, and validate (or invalidate) product concepts.
• Set technical direction for the hardest frontend problems: editor and canvas architecture, real-time collaboration, design-system inference, rich import/export, and performance as the product scales.
• Design and run experiments to test product hypotheses, from prototype tests with users to evaluations of generation quality.
• Own the quality bar for how the product looks and feels, treating latency, responsiveness, and reliability as first-class concerns.
• Strengthen the handoff to Claude Code so designs translate cleanly into production, and feed insights back to research about model effectiveness.
• Operate as a technical leader on a small team: shape the roadmap, make pragmatic architecture decisions, and mentor engineers as the team grows.

You may be a good fit if you:

• Have 8+ years of experience building full-stack applications with deep frontend strength, and a track record of zero-to-one work.
• Have set technical direction for a team or workstream and mentored other engineers, while staying hands-on in the code.
• Have strong design sensibility and a high bar for craft.
• Have strong technical skills across modern web stacks (React, TypeScript, Node.js, Python, Go), including complex client-side state, performance, APIs, databases, and cloud technologies.
• Are comfortable working in a fast-moving environment.
• Are deeply user-centric.
• Hold strong opinions loosely.
• Are a generalist who can move between frontend, backend, and product problems.

Strong candidates may also have:

• Extensive experience working with or building visual collaboration & creative tools (Figma, Canva, Adobe Creative Cloud).
• Hands-on experience with UI/UX for AI-powered applications.
• Background conducting user research, interviews, and usability testing.
• Deep frontend expertise in canvas rendering, complex editor UIs, design-system tooling, or graphics work (WebGL/shaders, 3D, animation).
• Experience with real-time, collaborative applications (multiplayer editing, WebSockets).

Candidates need not have:

• 100% of the skills listed above.
• Formal certifications or education credentials.
• Direct machine learning or AI research experience.

Annual Salary: $320,000 - $485,000 USD

Logistics:

• Location-based hybrid policy: we expect all staff to be in one of our offices at least 25% of the time.
• Visa sponsorship: We do sponsor visas. We can't successfully sponsor visas for every role and every candidate, but if we make you an offer we will make every reasonable effort to get you a visa.