Senior AI Infrastructure Engineer — Network Deployment

Nava; Bengaluru; On-site; INR 2,600,000 - 4,200,000; Full time

Job summary

Nava seeks a hands-on Network Deployment engineer to build and validate high-speed GPU networking for production AI clusters. You will configure InfiniBand/RoCE, Ethernet leaf-spine, and storage networks, and perform validation under load.

Ideal candidates have 5–8 years in data center networking, with strong BGP/L2/L3, EVPN/VXLAN, and automation with Python/Ansible. GPU-datacenter exposure is highly valued, along with RDMA and InfiniBand experience.

Qualifications

• 5–8 years in hands-on data center networking.
• Strong BGP, L2/L3, EVPN/VXLAN and high-speed switching knowledge.
• Linux debugging and Python/Ansible automation.
• Experience with Arista, Cisco, NVIDIA Networking or similar platforms.
• Within that, hands-on GPU / AI-datacenter exposure is valued.
• RDMA/RoCEv2, InfiniBand, PFC/ECN and GPU cluster testing.

Responsibilities

• Configure InfiniBand/RoCE, Ethernet leaf-spine, storage and management networking.
• Perform optics, cable, routing, topology and link-state validation.
• Test RDMA reachability, bandwidth, latency and fabric congestion under load.
• Automate switch provisioning, preflight tests and configuration checks.
• Investigate hardware, routing and network fabric issues with compute/storage teams.
• Produce acceptance evidence, as-built documentation and operational handover.
• Configure and tune PFC/ECN for lossless RoCEv2 and validate topology and routing.
• Use UFM (InfiniBand) and NetQ / Spectrum-X telemetry to verify PFC-pause is near zero and GPUDirect RDMA / BlueField DPU paths perform at line rate.

Skills

BGP

L2/L3

EVPN/VXLAN

High-speed switching

GPU Datacenter

RDMA/RoCEv2

Tools

Arista

Cisco

NVIDIA Networking

Python

Ansible

Linux debugging

Job description

About Nava

Nava is a neocloud company purpose-built for the AI era. We design, deploy, and operate large-scale GPU infrastructure and deliver inference-as-a-service to teams building the next generation of AI products. Our platform runs on NVIDIA GPU systems, high-performance RDMA fabrics, and a fully automated, software-defined operations model. Every engineer at Nava works close to the metal, on infrastructure built to keep GPUs saturated and models serving.

About Nava

Nava is a neocloud company purpose-built for the AI era. We design, deploy, and operate large-scale GPU infrastructure and deliver inference-as-a-service to teams building the next generation of AI products. Our platform runs on NVIDIA GPU systems, high-performance RDMA fabrics, and a fully automated, software-defined operations model. Every engineer at Nava works close to the metal, on infrastructure built to keep GPUs saturated and models serving.

About The Role

The Network Deployment team builds and validates the high-speed switching and GPU networking that connects Nava's production AI clusters. This is a hands-on engineering role.

What You'll Do

• Configure InfiniBand/RoCE, Ethernet leaf-spine, storage and management networking.
• Perform optics, cable, routing, topology and link-state validation.
• Test RDMA reachability, bandwidth, latency and fabric congestion under load.
• Automate switch provisioning, preflight tests and configuration checks.
• Investigate hardware, routing and network fabric issues with compute/storage teams.
• Produce acceptance evidence, as-built documentation and operational handover.
• Configure and tune PFC/ECN for lossless RoCEv2 (or InfiniBand credit flow), validate rail-optimized topology and adaptive routing, and run NCCL bus-bandwidth acceptance tests.
• Use UFM (InfiniBand) and NetQ / Spectrum-X telemetry to verify PFC-pause is near zero and GPUDirect RDMA / BlueField DPU paths perform at line rate.

What You'll Need

• 5-8 years in hands-on data center networking.
• Strong BGP, L2/L3, EVPN/VXLAN and high-speed switching knowledge.
• Linux debugging and Python/Ansible automation.
• Experience with Arista, Cisco, NVIDIA Networking or similar platforms.
• Within that, hands-on GPU / AI-datacenter exposure is what we weigh most - we value demonstrated depth over years, as the field is still emerging.
• RDMA/RoCEv2, InfiniBand, PFC/ECN and GPU cluster testing.

Skills: cluster,networking,infiniband,routing,infrastructure,rdma