Platform Engineer*in KI-Infrastruktur (w/m/d)

Job summary

The role develops a sovereign, multi-tenant observability platform and integrates customized language models into a federated Kubernetes inference infrastructure for research. Responsibilities include investigating AI-assisted operations and security analysis using MCP interfaces and agent-based approaches, improving GPU efficiency through techniques such as scale-to-zero and model-weight caching, and publishing findings as research and reference architectures.

Benefits

Qualifications

• Completed scientific university degree in Computer Science, Business Informatics, or a related field
• Strong knowledge of container technologies and Kubernetes
• Experience with monitoring and observability tools
• Basic understanding of IT security in cloud environments
• Interest in scientific work and publishing results
• Good German and English language skills
• Experience operating language models such as vLLM, using MCP or agent-based AI systems, or working with HPC environments is desirable
• Independent, structured working style and good communication skills

Responsibilities

• Design, evaluate, and build a sovereign, multi-tenant observability platform based on open-source software
• Integrate customized language models into a federated Kubernetes inference infrastructure
• Research AI-assisted operations and security analysis using language models, MCP interfaces, and agent-based approaches
• Investigate efficient GPU utilization, including scale-to-zero, GPU virtualization, and model-weight caching
• Publish findings in academic venues and develop reference architectures and best practices

Skills

Degrees

Languages

Work schedule

Industry

Contract length

Relocation