Runpod is the AI Developer Cloud, empowering over one million developers to experiment, train, and scale AI. They are seeking a Cloud Infrastructure Security Engineer to safeguard their GPU cloud infrastructure through systems-level security engineering and robust isolation architectures.
Responsibilities:
- Design and implement robust workload and network isolation architectures for RunPod's multitenant GPU bare-metal and virtualized environments
- Harden Linux kernel configurations, container runtimes (e.g., Docker, containerd), and orchestration layers (e.g., Kubernetes) against breakouts and privilege escalation
- Conduct deep-dive security assessments and penetration testing specifically targeting our hypervisor, network stack, and hardware interfaces
- Write code (primarily C, Go, or Rust) to implement custom security controls, telemetry, and fixes at the OS and infrastructure level
- Evaluate and mitigate security considerations specific to GPU architecture, PCIe pass-through, and shared memory spaces
- Serve as the technical escalation point for infrastructure-level security incidents, developing forensic capabilities for ephemeral container environments
Requirements:
- 5+ years of experience in infrastructure or systems-level security engineering
- Extensive knowledge of Linux kernel internals (cgroups, namespaces, eBPF, SELinux/AppArmor)
- Deep understanding of virtualization technologies (KVM, QEMU) and workload/network isolation techniques in multitenant environments
- Strong systems-level programming skills in C, Go, Rust, or Python
- Familiarity with GPU architecture and hardware-level security considerations
- Experience in securing bare-metal cloud infrastructure and mitigating lower-level CVEs
- Contributions to open-source systems security projects or virtualization research
- Experience writing or deploying eBPF-based security tooling
- Deep knowledge of low-level networking protocols and virtualized network security