Mirantis
Mirantis
Mirantis develops cloud-native infrastructure and container management software for organizations building, securing, and operating Kubernetes environments at scale. Its portfolio includes Mirantis Kubernetes Engine, Mirantis OpenStack for Kubernetes, Mirantis Container Cloud, Mirantis Container Runtime, and Mirantis Secure Registry, supporting cluster operations and software supply-chain security. The company also contributes to the open-source ecosystem through projects and tools such as Lens Desktop, a Kubernetes IDE, alongside enterprise technical support. Mirantis works with organizations in consulting, healthcare, logistics, public services, financial services, and technology-focused industries adopting modern cloud infrastructure.

Senior AI Infrastructure Networking Solutions Architect

Design GPU cluster networking for Mirantis’s Kubernetes-native AI infrastructure platform. Build reference architectures, automation, and customer proofs of concept for k0rdent AI.

Description

  • Lead network solution design for Mirantis’s k0rdent AI platform for GPU clouds and AI factories.
  • Design interconnects for large GPU clusters, tenant isolation, and networking across containers, virtual machines, and bare-metal nodes.
  • Develop and publish reference architectures for deployments ranging from a single rack to clusters with thousands of GPUs.
  • Define InfiniBand and RoCEv2 Ethernet compute fabrics, including rail-optimized and fat-tree or Clos designs.
  • Design front-end, storage, and management networks, plus connections to customer data centers, public clouds, and hybrid environments.
  • Specify multi-tenant isolation with InfiniBand partitions, VRFs, EVPN-VXLAN, and Kubernetes network policies.
  • Document scaling limits and trade-offs in cost, performance, operability, and vendor dependence.
  • Define Linux host networking and how workloads access NICs, DPUs, and SuperNICs.
  • Design Kubernetes networking for AI workloads using CNIs, Multus, SR-IOV, RDMA plugins, NVIDIA Network Operator, and DRA.
  • Design KubeVirt networking for GPU and RDMA traffic.
  • Research emerging AI networking technologies and standards, prototype designs, and compare results with current practices.
  • Publish research notes, design proposals, papers, blog posts, and talks.
  • Build and run proofs of concept with customers and partners, as well as in Mirantis labs.
  • Develop automation and tools with Python, Go, Bash, Ansible, Helm, Kubernetes manifests, and Terraform.
  • Test and benchmark fabrics and host configurations with tools such as NCCL tests, perftest, and ib_write_bw.
  • Serve as the networking expert in customer discovery, design reviews, and architecture workshops.
  • Work with hardware and networking partners, Solutions Architects, Partner Management, Engineering, Product, and customers.
  • Present at industry events, webinars, and partner summits.
  • Lead practical technical workshops and create technical enablement materials.
  • Share research findings with Product and Engineering and help guide the k0rdent AI roadmap.

Requirements

  • Hold a bachelor’s degree in Computer Science, Electrical Engineering, Telecommunications, or a related field, or have equivalent practical experience.
  • Bring at least eight years of network engineering or architecture experience.
  • Have at least three years of experience with data center, HPC, or cloud infrastructure networking.
  • Have customer-facing experience as a solutions architect, pre-sales engineer, consultant, or technical lead.
  • Bring data center network design expertise, including spine-leaf and Clos topologies, rail-optimized GPU fabrics, oversubscription, ECMP, and scaling to thousands of nodes.
  • Have InfiniBand expertise covering subnet management, partitioning, adaptive routing, and NCCL/RDMA traffic behavior.
  • Have RoCEv2 Ethernet expertise covering PFC, ECN, DCQCN, QoS, MTU, and buffer tuning.
  • Understand routing and overlays, including BGP, EVPN-VXLAN, and VRFs for multi-tenancy.
  • Have experience with hybrid and multi-cloud connectivity, interconnects, VPNs, cloud networking constructs, IP address planning, and DNS planning.
  • Know Linux networking, NIC drivers, PCI passthrough, SR-IOV, IOMMU, NUMA affinity, iproute2, ethtool, and devlink.
  • Know Kubernetes networking, including CNI plugins, Multus, SR-IOV and RDMA device plugins, network policy, and service exposure.
  • Have experience with KubeVirt and VM networking, passthrough, and SR-IOV.
  • Be able to program or script in at least one language; Python or Go is preferred.
  • Be comfortable working with Git, CI, and infrastructure as code.
  • Have excellent spoken and written English.
  • Be comfortable presenting to large audiences and leading hands-on workshops.
  • Be able to adapt communication for network engineers and customer executives.
  • Have experience working in an international, distributed company across time zones and work cultures.
  • Be able to travel up to 25%.
  • Equivalent practical experience may replace the bachelor’s degree requirement.

Benefits

  • Compensation and benefits are determined by the local market and employment arrangement.
  • Remote work arrangement.
  • Travel opportunities for customer engagements, partner meetings, lab work, and industry events, with travel of up to 25%.
  • Work with a distributed team.
  • Work directly with leading GPU cloud operators, NeoClouds, sovereign clouds, and AI-focused enterprises.
  • Help shape the product narrative and influence go-to-market success.

Related Jobs

dentsu Austria

Analyst Relations & Insights Manager (Hybrid, United States)

dentsu Austria
51 – 200 Employees
EnterpriseMarketingMedia

Manage analyst relations programs, evaluation submissions, and research insights for Merkle. Coordinate client stories, internal communications, and engagement with leading analyst firms.

Open