Key Responsibilities
- Build and deliver custom environments with excellent GPU performance for customer workloads.
- Leverage AI to an extreme level to automate provisioning, alerting and recovery.
- Provision and configure dedicated Kubernetes clusters tailored to customer requirements.
- Design and implement overlay networking (VLAN, VXLAN) and routing configurations (ECMP, BGP) and tunnels (strongSwan, IPSEC) for tenant isolation and performance.
- Build and maintain Linux images.
- Set up network monitoring and diagnostics for customer environments.
- Automate the end-to-end lifecycle of customer compute environments: creation, configuration, validation, and teardown.
Requirements
- 5+ years experience with Linux virtualization: KVM/QEMU, libvirt, VFIO device passthrough, hugepages, NUMA, CPU pinning.
- Strong networking fundamentals: VXLAN, VLAN, ECMP, BGP, ARP, and the ability to debug packet-level issues (tcpdump, Wireshark).
- Production experience building and operating Kubernetes clusters on bare metal (MetalLB).
- Proficiency with Linux image building and OS provisioning (kickstart, cloud-init, PXE/iPXE).
- Proficiency in Python, Bash, Ansible and Terraform.
- Deep experience with NVIDIA GPUs: drivers, MIG, container runtimes (nvidia-container-toolkit), InfiniBand, RDMA/RoCEv2 and GPUDirect for high-performance AI networking.
- Excellent communication and ability to drive technical decisions across teams.
- Self-starter who executes quickly, takes ownership, and constantly seeks improvement.
Nice-to-Haves
- Experience with SR-IOV, DPDK, or other high-performance networking technologies.
- Experience with shared network storage (Ceph, Lustre, Weka).
- Experience with network automation tools (Netbox, Nautobot, Nornir).
Compensation
$180,000-250,000 plus equity + benefits (This range encompasses 2 levels Senior and Staff).