AI Cloud Network Architect

You will design and plan network architecture for AI data centers, internal networks, data center interconnects, and global backbones. You will develop GPU cluster interconnects, underlay and overlay designs, congestion-control standards, architecture documentation, topology diagrams, configuration standards, and improvement roadmaps for large-scale network environments.

Responsibilities

• Translate business and cloud platform requirements into network architecture designs.
• Plan IP addressing VLAN VXLAN and bandwidth resources.
• Design AI data center networks including GPU interconnects front-end networks and management networks.
• Design InfiniBand and RoCE cluster interconnects and Spine-Leaf or Clos topologies.
• Integrate underlay networks with SDN and VXLAN or EVPN overlays.
• Tune PFC ECN and in-band network telemetry.
• Produce HLD and LLD documents topology diagrams configuration standards and SOPs.
• Identify network architecture risks and drive improvement roadmaps to completion.
• Engage vendors ISPs and colocation partners on network planning and implementation.

Requirements

• Bachelor's degree or above in computer science telecommunications electronics or a related field.
• 8+ years of network architecture design and planning experience.
• Experience at a hyperscale internet company cloud provider or multinational backbone operator.
• Expertise in InfiniBand and RoCEv2 architecture.
• Knowledge of rail-optimized high-performance network topologies.
• Experience planning or deploying 10000- to 100000-GPU-scale networks is a strong plus.
• Experience with BGP OSPF ECMP VXLAN EVPN and Segment Routing.
• Knowledge of SR-MPLS or SRv6.
• Knowledge of optical transport DWDM and global backbone peering or transit strategies.
• Strong cross-functional and cross-cultural communication skills.