Munich HQ (hybrid) | Full-time | Experienced Hires

<aside> 💡

Zalion is on a mission to create the operating system for autonomous procurement. We are moving beyond "Chat with PDF" to build agentic AI that operates deep within the enterprise, negotiating via email, navigating legacy ERPs, and managing complex Source to Pay workflows to deliver real impact.

Join us early and help define how enterprise AI is done right.

</aside>

🪛 How we operate

• You’ll work closely with the founders and early customers to understand real procurement workflows and turn them into robust, AI-native product experiences.
• We are fostering the use of agentic coding, reviewing support, security checks etc. and expect you to shape the responsible use of it within the engineering organization.
• We are straight shooters. Clear, respectful and proactive communication is the key driver in our culture.

🚀 What You’ll be responsible for

You will:

• Own and ship agent-powered features end-to-end – from idea and prototype to production rollout and iteration.
• Create the Zalion Procurement Harness and solve hard problems around context management, latency, robustness, and guardrails – making agents debuggable and trustworthy, not magic.
• Design, build, and deploy AI agents in production that interact with real systems (ERPs, email, supplier portals) and deliver measurable outcomes.
• Work with both structured and unstructured data (RFQs, emails, quotes, contracts) to power memory, retrieval, and decision-making.
• Build scalable memory, reasoning, and orchestration pipelines using tools like LangChain/LangGraph and evaluation frameworks such as Langfuse.
• Contribute to our cloud infrastructure for scalable agentic workloads (observability, CI/CD, reliability, cost-awareness).

🛠️ Tech Stack You’ll Work With

• Python (LangChain, LangGraph, FastAPI)
• Langfuse, MLFlow, Signoz, GitHub CI/CD, PostHog, Temporal
• Anthropic, OpenAI, Gemini, Mistral Inference