• Design, build, and ship Personio's internal products and tools end-to-end: complex agents, cross-functional automations, and the applications and interfaces around them
• Make the architecture calls: choose the right patterns, manage complexity, and make pragmatic trade-offs between reliability, cost, and delivery speed
• 
  

Productionise and maintain

• Define what "done" means for an internal AI product: secure auth, error handling, monitoring, rollback, and documentation, then build to that bar
• Own systems over their full lifetime: maintain, monitor quality with Personio's eval platform, catch degradation, and keep them reliable long after launch
• 
  

Contribute to the AI platform

• Build on and extend Personio's internal AI platform: agent orchestration, the eval platform, context layer, and connector library
• Feed production experience back into the platform: the use cases you build surface the requirements, and you collaborate with AI Platform engineers to evolve shared infrastructure and the skills library so what you build compounds across the team
• 
  

Engineering standards

• Contribute to Personio's internal AI engineering standards: what production-grade solutions look like, how they are evaluated, deployed, and maintained
• Document solutions clearly so they can be maintained, extended, and handed over without you in the room
• 
  

Role Requirements: What you need to succeed

• 4+ years building and shipping production software, working independently
• Strong Python skills — our primary backend language — across APIs, services, and data integration
• Hands-on experience building production AI products: agentic workflows and RAG pipelines with LLM APIs, agent frameworks (e.g. LangGraph), and MCP