Physical AI Staff AI Systems Software Engineer

I’m working with a very well-funded, high-ambition Physical AI company in San Francisco that is building models capable of designing real-world physical products. They’re hiring a Software Engineer / Agentic Systems Engineer to help build the orchestration and workflow layer around these systems. This is a strong fit for someone who has worked on agentic workflows, multi-agent orchestration, and developer tooling such as Claude Code or similar coding-agent environments. This is an opportunity to join a category-defining company working at the intersection of AI, software infrastructure, and physical design.You would be responsible for building the systems that allow AI models to reliably plan, coordinate, use tools, and execute complex design tasks across CAD-driven workflows.Key Responsibilities of the Role: Building orchestration workflows for agentic systemsDesigning and scaling multi-agent systems for decomposition, planning, checking, and refinementWorking with Claude Code and related coding-agent toolingDeveloping infrastructure for long-running, stateful, tool-using model workflowsImproving reliability, latency, observability, and evaluation across agent pipelinesCollaborating closely with research and product teams to productionise cutting-edge Physical AI systems Key Qualifications Needed: A solid background in software engineering, backend systems, or infrastructureExperience with LLM orchestration, agentic systems, or multi-agent frameworksExperience in building systems utilizing multi modal data Familiarity with tool-using AI systems and complex workflow executionHands-on exposure to Claude Code or similar AI-assisted coding / agent platformsStrong Python skills and experience building production-grade systemsThe ability to operate in a fast-moving, high-ownership environment Required Skills:A solid background in software engineering, backend systems, or infrastructureExperience with LLM orchestration, agentic systems, or multi-agent frameworksFamiliarity with tool-using AI systems and complex workflow executionHands-on exposure to Claude Code or similar AI-assisted coding / agent platformsStrong Python skills and experience building production-grade systemsThe ability to operate in a fast-moving, high-ownership environmentPay range and compensation packageCompensation is highly competitive, with base salary up to $350,000 plus equity. The role is hybrid in San Francisco.