Staff Software Engineer, Labs: Applied AI

About the role

At Anthropic, we're building AI systems that are safe, beneficial, and transformative. Our mission is to develop AI that benefits humanity, and we believe the most powerful capabilities emerge when we thoughtfully bridge the gap between research breakthroughs and real-world applications.

Applied AI is one of the newest explorations within Anthropic Labs, the internal accelerator behind Claude Code, MCP, and Claude Design. Most of the world's work happens far from a code editor, and the people doing it have barely begun to feel what frontier AI can do. We believe Claude has a transformative role to play here — and we're at the earliest stage of exploring what that could look like. The engineers who join now will define where it goes.

We're looking for versatile, entrepreneurial engineers who are energized by building for users unlike themselves. In this role, you'll take frontier AI capabilities and turn them into applications that professionals in less software-native roles can pick up and trust — rapidly building and testing new experiences, partnering directly with researchers, domain experts, and users, and generating the insights that shape where this exploration goes next. You'll need to be comfortable with ambiguity, willing to kill your own projects when the data says to, and energized by the pace of building in uncharted territory.

Responsibilities

• Rapidly prototype full-stack applications that bring frontier AI into workflows that have never been software-first, shipping early and often to maximize learning

• Immerse yourself in unfamiliar domains: sit with users, learn how their work actually gets done, and encode that understanding into products, evaluations, and workflows

• Collaborate closely with research teams to understand new model capabilities and translate them into tools that non-technical professionals reach for first

• Work directly with internal teams and external partners across industries to gather feedback, iterate quickly, and validate (or invalidate) product concepts

• Design and run structured experiments to test hypotheses, balancing creative exploration with rigorous evaluation

• Generate documentation and insights to guide successful prototypes toward full product teams

• Provide feedback to research teams about model effectiveness in real-world, domain-heavy settings and where capabilities can improve

• Flexibly contribute across Labs initiatives based on organizational priorities and emerging opportunities — context from one project should inform the next

You may be a good fit if you

• Have 8+ years of experience building full-stack applications, with a track record of zero-to-one work in startup or startup-like environments

• Are deeply curious about how other industries work, and enjoy translating messy, real-world workflows into simple software

• Thrive in ambiguity and are energized (not anxious) by uncertainty — you're comfortable working on projects that might not exist in three months

• Have a hacker mentality: high agency, bias toward shipping, comfort with technical debt when it's the right tradeoff

• Are deeply user-centric — you validate ideas with actual users before over-investing and talk about problems before solutions

• Can articulate learnings from failed or killed