Compute Infrastructure Engineer – NeuraGym (Mensch)

Job summary

The Compute Infrastructure Engineer designs, builds, and operates scalable compute and storage infrastructure for multimodal robotics data, enabling reliable, high-performance, reproducible recording and processing. The role connects servers, workstations, and robot edge devices; manages virtualization and Kubernetes environments with GPU scheduling; and develops distributed data pipelines, networking, and infrastructure automation for embodied AI systems.

Qualifications

• At least 4 years of experience in system administration, infrastructure engineering, DevOps, or platform engineering.
• Experience planning, designing, and operating server infrastructure, including sizing, rack assembly, hardware selection, and procurement.
• Practical experience with hypervisors and virtualization, such as Proxmox, KVM, or VMware, including storage and network integration.
• Knowledge of container technologies and orchestration, such as Docker and Kubernetes, ideally including GPU workloads.
• Understanding of distributing, synchronizing, and monitoring processes, data, and compute workloads across multiple machines.
• Proficiency with Linux, shell scripting, and at least one scripting language, such as Python or Bash.
• Experience with infrastructure as code and automation, such as Ansible or Terraform, and CI/CD tools such as GitLab CI.
• Structured, independent working style and ability to build infrastructure collaboratively with an interdisciplinary team.
• Experience with monitoring and logging stacks such as Prometheus and Grafana is an advantage.

Responsibilities

• Plan and size server, GPU, storage, and network infrastructure, including rack layout, power, cooling, and hardware procurement.
• Build and operate virtualization environments and provide development, test, and production systems.
• Design and operate Kubernetes or comparable clusters for data preprocessing, including GPU scheduling, monitoring, and logging.
• Distribute compute- and data-intensive workloads across robot edge devices, local servers, and central storage, and identify and resolve bottlenecks.
• Design network infrastructure, including VLANs, switching, bandwidth planning for camera and sensor streams, firewall rules, and secure integration with enterprise IT.
• Build storage and data pipelines for large datasets, ensuring reliable capture, storage, versioning, and availability for model training.
• Automate provisioning, deployment, and operations with infrastructure-as-code and CI/CD tools, and document the system landscape.
• Resolve infrastructure performance, stability, and security issues and develop backup, recovery, and access-control plans.

Skills

Degrees

Industry

Relocation