Neo Cloud | AI Cloud Network Engineering, MOT

About the Role

Key Responsibilities

• Design, deploy, and operate the data center network for our GPU clusters: leaf-spine (Clos) fabrics running BGP, EVPN/VXLAN, and ECMP across the front-end, storage, and out-of-band management networks.
• Build multi-tenant isolation into the fabric with VRFs, VLANs/VNIs, overlays, and policy-based segmentation, so every customer gets a private, high-performance network.
• Bring up new clusters and sites: turn designs into configurations, validate cabling and optics, and run burn-in and acceptance testing with our data center and hardware partners.
• Partner with compute and storage engineers on the RDMA fabrics (RoCE/InfiniBand) that move data between GPUs and storage, including QoS and congestion control.

Network Functions

• Design and operate the network functions that sit at the edge of and between tenants: NAT/CGNAT, stateful firewalls and ACLs, load balancers, VPN gateways, DDoS protection, DNS/DHCP, and IPAM.
• Choose between hardware appliance, virtual (VNF), and host/DPU-based implementations, weighing throughput, state, failover, and cost.
• Build security into the network: zero-trust segmentation, management-plane hardening, and audited, reviewable changes.

WAN, Edge, and Interconnect

• Connect our data centers to the internet: IP transit and peering with ISPs and internet exchanges, BGP routing policy, RPKI and route filtering, and IP address and ASN management.
• Build private connections to other clouds (AWS Direct Connect, Google Cloud Interconnect, Azure ExpressRoute, OCI FastConnect) and to customer networks, with IPsec VPN where a dedicated link doesn't make sense.
• Design and run the backbone between our own sites (wavelengths, dark fiber, MPLS or segment routing), with traffic engineering and capacity planning as we add regions.
• Select and manage carriers, colocation providers, and network hardware vendors: circuits, cross-connects, hardware qualification, and RMAs.

Automation and Operations

• Run the network as code: a source of truth, templated configurations, Git-based review, CI validation, and automated rollout and rollback.
• Work with our AI agents and digital twin to validate changes before they reach production and to let agents deploy and monitor the network safely.
• Instrument the network with streaming telemetry, sFlow/IPFIX, and synthetic probes; build alerting that catches problems before customers do.
• Take end-to-end ownership of the network in production: participate in on-call, lead incident response and root-cause analysis, and drive long-term reliability and performance improvements.
• Lead capacity planning for rapid growth, from port counts and optics to transit and interconnect bandwidth.

Technical Leadership

• Strong bias to action and resolution of technical decisions.
• Set network standards, change-management practices, and runbooks; author and review design documents for significant changes.
• Partner with storage, compute, and security engineers to deliver complete AI infrastructure.
• Mentor engineers and raise the bar through design review, config review, and hands-on collaboration.

Ideal Candidate

• 8+ years designing, building, and operating large-scale production networks at a cloud provider, neocloud, hyperscaler, service provider, or HPC site.
• Deep TCP/IP knowledge and hands-on expertise with BGP, IS-IS/OSPF, EVPN/VXLAN, ECMP, and QoS in leaf-spine data center fabrics, including IPv6.
• Hands-on experience configuring and operating network functions in production: NAT/CGNAT, stateful firewalls, load balancers, IPsec VPN, DNS/DHCP, and IPAM.
• Experience connecting networks to the outside world: IP transit and peering, internet exchanges, BGP policy and route security, and private interconnects to AWS, Google Cloud, Azure, or OCI.
• Experience with multi-site networks: backbone and WAN design, MPLS or segment routing, and routing across multiple data centers.
• Production experience with several vendors and network operating systems, such as Arista EOS, Juniper Junos, Cisco NX-OS/IOS-XR, NVIDIA Cumulus, or SONiC.
• Strong Linux skills, including the Linux networking stack, and network automation in Python with Git-based workflows and a configuration management tool (Ansible, Salt, or similar).
• Strong troubleshooting from packet captures to routing tables to vendor bugs, using tools like tcpdump, Wireshark, and mtr.
• Proven experience owning networks in production, including on-call, incident response, and driving systemic reliability improvements.
• Excellent written and verbal communication skills.

Nice to have

• Hands-on experience with RoCEv2 and/or InfiniBand fabrics for GPU clusters, and an understanding of AI training and inference traffic patterns.
• Optical transport and DWDM experience, and qualifying high-speed optics (400G/800G).
• Experience running network functions on DPUs/SmartNICs (e.g., NVIDIA BlueField) or software data planes (eBPF/XDP, DPDK, VPP).
• Software-defined networking and cloud overlays; Kubernetes networking and CNIs such as Cilium or Calico.
• Network observability tooling such as gNMI streaming telemetry, Prometheus, Grafana, or OpenTelemetry.
• Experience bringing up the network for a greenfield data center or a new cloud region.
• Experience validating and safely deploying AI-generated automation and configuration.
• Certifications such as CCIE or JNCIE.