Senior Agent Platform Engineer

Proxima Fusion — Senior Agent Platform Engineer

Proxima Fusion is Europe’s fastest-growing fusion company and the continent’s best-funded fusion player. We are building hardware and infrastructure to deliver the world’s first commercial stellarator fusion power plant. Our work combines stellarator optimization, advanced computation, machine learning, and high-temperature superconducting magnets to unlock higher-performance designs.

Location: Munich, on-site

About the role: You will own the design, delivery, and operation of our agent platform, turning autonomous agents into reliable, production-grade capabilities for physicists and engineers. This includes durable sessions, sandboxed execution environments, long-running tasks, memory management, and tool integration. You will co-design the integration layer connecting agents to internal and external systems, ensure safe and governable autonomy, and implement observability and reliability across Google Cloud Platform (GCP).

What you will do

• Own the design, delivery, and operation of our agent platform: including runtime primitives, durable sessions, sandboxed code execution, long-running tasks, memory management, and tool integration.
• Stabilize a rapidly evolving agent ecosystem by evaluating and selecting supported primitives vs experiments.
• Co-design the integration layer connecting agents to internal/external systems and standardize tools, skills, and context exposure.
• Ensure safety and governance: human-in-the-loop approval, capability boundaries, audit logging, cost controls, and defense against prompt injection.
• Develop evaluation and observability for non-deterministic systems: tracing, quality evaluations, and cost/latency visibility.
• Operate deployments on Google Cloud with reliability and security considerations.
• Collaborate with demanding technical users and shape the platform roadmap with the team lead.

Who you are

• 5+ years of software engineering experience with Python and distributed backend systems.
• Experience architecting, building, and operating production-grade agentic or LLM systems beyond prototypes.
• Hands-on knowledge of agent ecosystems (MCP, agent SDKs/CLIs, multi-agent orchestration, sandboxed code execution).
• Experience running production workloads on Kubernetes and a major cloud (ideally GCP).
• Strong emphasis on evaluation and observability as platform features; security-minded design.
• Love for platform thinking, internal products, and clear communication with expert users.
• Experience with Infrastructure as Code (Terraform) and CI/CD.

Nice to have

• Model gateways and routing, HPC or scientific computing exposure, and open-source contributions to AI/agent tooling.

Interview process

• Recruiter Interview (30-60 min)
• Technical Screening (30 min)
• Technical Panel (3x60 min)
• CEO call (30 min)

Join us to help shape how AI and automation accelerate fusion research and engineering. This role sits at L2 of our framework; inquire during recruitment for details.