Sovereign AI Platform Engineer - T Cloud Public (REF5674Y)

As Hungary’s most attractive employer in 2025 (according to Randstad’s representative survey), Deutsche Telekom IT Solutions is a subsidiary of the Deutsche Telekom Group. The company provides a wide portfolio of IT and telecommunications services with more than 5300 employees. We have hundreds of large customers, corporations in Germany and in other European countries. DT-ITS recieved the Best in Educational Cooperation award from HIPA in 2019, acknowledged as the the Most Ethical Multinational Company in 2019. The company continuously develops its four sites in Budapest, Debrecen, Pécs and Szeged and is looking for skilled IT professionals to join its team. Mission Design, build, and operate a sovereign AI toolchain used in isolated or tightly controlled environments, including model serving, retrieval, orchestration, observability, and secure platform operations for enterprise AI engineering workloads Role focus This role centers on open-source or open-weight LLM stacks, air-gapped or isolated deployment models, Kubernetes-based platform engineering, GPU-enabled environments, and auditable AI operations suitable for sovereignty-sensitive programs Key responsibilities Design and run Kubernetes environments optimized for AI inference, retrieval, experimentation, and agent execution in secure or isolated settings Deploy and operate open-source or open-weight model stacks, model gateways, vector databases, and supporting platform components Build reproducible platform automation using Infrastructure as Code and GitOps approaches for stable, auditable delivery Manage local registries, package mirrors, secrets, access controls, storage, networking, and observability in environments with limited or no public cloud dependency. Optimize GPU, compute, and storage usage for reliable AI workloads while maintaining security and data sovereignty requirements Examples of market tools, models, and platform components expected Inference and local serving stacks such as vLLM, Olla