Applied AI Engineer, Startups (Physical AI)

About the role

As an Applied AI Engineer on the Startups team at Anthropic, you will be a trusted technical advisor to the top AI-native startups building in the physical world, across robotics, semiconductors, aerospace, manufacturing, and civil infrastructure, as they grow from early product to scale. Some of these companies build the hardware itself and want Claude woven into how they design, test, and operate it, and others build the software that engineers everywhere use to design, simulate, test, and scale, and this role works with both. You'll lead deep technical engagements, partnering directly with startup software and hardware engineering teams to help them build their products on Claude and run their own engineering with it, advising on implementation and agent design as they build what is newly possible with AI in the physical world. You will bridge the gap between what Claude can do and what it takes to build real things, working alongside some of the strongest teams in the field on problems that matter and helping mechanical, electrical, and process engineers get the kind of step change that software engineering has had from agentic coding with tools like Claude Code.

Working closely with Applied AI Architects, Account Executives, and Anthropic's Product, Engineering, and Research teams, you'll help startups build at the edge of what is possible with Claude. You'll leverage your AI engineering expertise and your engineering domain knowledge to develop custom evaluation frameworks, design scalable architectures, and create the technical resources that enable physical AI startups to succeed with Claude.

Responsibilities

• Serve as a deep technical advisor to high-potential startups building in the physical world, whether Claude sits inside the product they sell or inside the way they design, test, and run their own hardware, working alongside them to push the boundaries of AI in engineering, manufacturing, operations, and more

• Work hands-on with startup software and hardware teams through pair programming and architecture reviews that accelerate their development, and understand their engineering workflows end to end, from design and simulation through test, production, and operations, so that your guidance fits how real hardware gets built

• Develop prototypes and technical documentation, including evaluation suites, benchmarks grounded in real engineering tasks, AI engineering techniques, and architecture diagrams, that enable startups to build, operate, and scale with Claude

• Create reference implementations and sample code that many companies can adopt without starting from scratch, for example integrating Claude with CAD, simulation, design tools, or industrial automation

• Collaborate closely with Applied AI Architects to maintain context and continuity across customer engagements

• Identify patterns across engagements, including what is hard about applying AI to physical systems such as reading drawings, 3D geometry, board layout and routing, GD&T, latency and on-premise constraints, and physical safety, then build repeatable evaluations and solutions and contribute insights back to Product, Engineering, Research, and the broader Applied AI team

• Create technical content for startup audiences including documentation, tutorials, and sample code

• Foster community engagement through hackathons, webinars, technical office hours, and startup-focused events

• Travel to customer sites for workshops, implementation guidance, and relationship building

You may be a good fit if you have

• Deep engineering experience in a physical discipline such as mechanical, electrical, aerospace, robotics, semiconductor, manufacturing, or civil engineering, gained by designing, building, or operating real hardware, along with strong opinions of your own about how to bridge the gap between what AI can do today and how physical things get built

• 4+ years of technical experience as a software engineer, ML or research engineer, robotics or controls engineer, hardware or systems engineer, or technical founder, with either production experience building LLM-powered applications, including prompting, context engineering, agent architectures, evaluation frameworks, and deployment at scale, or a track record of successfully applying AI to real engineering work in design, simulation, test, or production

• Experience with startups or high-growth companies and a love of working with founders and small teams that move quickly

• Strong technical communication skills, with the ability to translate complex AI concepts into architectural decisions and actionable implementation plans for both software and hardware engineering audiences and to run technical workshops, hackathons, or developer-focused events

• A collaborative approach to working across teams, the range of a generalist who can context-switch from robots to chips to rockets to factories and get up to speed on a new domain quickly, and a passion for making powerful technology safe and beneficial