Senior Frontend & Agent Engineer (f/m/d): Joule Work Desktop

We help the world run better At SAP, we keep it simple: you bring your best to us, and we'll bring out the best in you. We're builders touching over 20 industries and 80% of global commerce, and we need your unique talents to help shape what's next. The work is challenging – but it matters. You'll find a place where you can be yourself, prioritize your wellbeing, and truly belong. What's in it for you? Constant learning, skill growth, great benefits, and a team that wants you to grow and succeed.    The Team You will join a small team of experts within SAP's AI engineering unit, focused on building Joule Work Desktop. The team operates with high autonomy, high ownership, and a bias toward shipping. You will work alongside engineers who treat AI not as an add-on but as a fundamental part of how they build software. The Product Joule Work Desktop is SAP's AI-powered desktop application for knowledge workers and business users. It is a local, agent-driven environment that puts an AI agent directly into the user's daily workflow. The product combines a UI layer built with TypeScript and React with an underlying agent harness that executes multi-step tasks, integrates with external services via MCP servers, and is extensible via a skills system. It demands equal strength in AI engineering and product-quality UI work. What You'll Do This is a hands-on individual contributor role. Your work will span two core areas: Agent Harness . You will extend, improve, and optimize the agent harness powering Joule Work Desktop: skill selection, tool use, MCP server connectivity, correctness, speed, and security. You will work on observability, explainability, and evaluation infrastructure that keeps the agent reliable and auditable. UI & Product Features . You will ship user-facing features in TypeScript and React, working across the application's interface layer. In both areas, the expectation is fast iteration and continuous shipping. The team is small by design. You use AI tooling to move at a pace that would otherwise require more people.   What you bring Required - TypeScript & React . Senior-level proficiency in building production UI applications. You write clean, maintainable TypeScript and are comfortable owning complex React codebases. -  Agent Engineering . Significant hands-on experience building, operating, or optimizing AI agents and/or agent harnesses. You understand how agents plan, execute, recover from errors, and interact with tools. You have worked on problems of agent correctness, latency, and security in practice, not only in theory. -  AI Systems Knowledge . You are fluent in the concepts that agent systems are built on: skills and tool-use patterns, MCP server integration, observability and tracing of agent behaviour, explainability of agent decisions, and evaluation frameworks for measuring agent quality. -  AI-Native Engineering Practice . You use AI to write, review, and ship code faster than would otherwise be possible. You have deep, daily experience with tools like Claude Code, Codex, Cursor, or equivalent AI-centric development environments. When we ask how you use AI in your work, we expect concrete answers: what you shipped, how long it took, what wouldn't have been feasible otherwise. We also expect opinions on how AI-native engineering should be done, not just that you've tried it. -  University Degree . In computer science, software engineering, or a closely related field. Preferred - Experience with  LangGraph, LangChain , or similar agentic orchestration frameworks. - Experience building  Electron  applications. - Experience with  Windows development  OS APIs, platform-specific quirks, and native modules. - Familiarity with  application security, sandboxing, or process isolation  understanding how to constrain untrusted code execution in desktop environments. - A track record of shipping AI-powered products to real users, not just prototypes or demos.