Senior AI Engineer, Agents and Platform

• Own the platform every anzel agent runs on and set its architecture
• Decide what to build, buy, and retire, documenting decisions for people and agents
• Build the harness that turns plain-language skills into reliable agents
• Develop tools, orchestration, long-running workflows, memory and context, retrieval, structured output, model routing, fallbacks, and rollback
• Design the tool layer, including tool assignment, versioning, access scoping, and per-customer logging
• Build safe agents and separate customer data in collaboration with the head of engineering
• Defend against prompt injection and record every agent action
• Run evaluations on every change and trace cost and execution time
• Maintain continuous delivery, review with coding agents doing the first pass, and releases that roll back quickly
• Lead by building, making clear decisions, and providing fast, kind reviews

• Eight or more years building and running production software, or the equivalent built on your own
• Real backend depth, including APIs, a relational database, queues, and long-running workflows
• Experience building language model products relied on by real users
• Experience with retrieval and agents beyond a demo
• Judgment to know when an agent is the wrong tool
• Experience designing evaluations that prove whether a change helped
• Daily use of coding agents, several at once, without shipping code nobody understands
• Security mindset from running systems that held customer data
• Something built that can be shown
• Professional English
• German is a plus

• Your own machine
• Your own choice of tools
• Highly competitive pay reflecting the seniority and responsibility of the role
• Startup pace supported by XU Group's experience and network
• Opportunity to own the platform of a new product with real customers from day one
• Small senior team with almost no bureaucracy