Network Engineer — GPU Infrastructure

Nava; Bengaluru; On-site; INR 900,000 - 1,300,000; Full time

Job summary

Nava in Bengaluru is seeking a capable Systems Engineer to design, deploy, and operate GPU infrastructure for AI workloads. You will manage bare‑metal GPU clusters and ensure fast, reliable training and inference at scale.

You will implement OS provisioning, CUDA drivers, GPU Operator, and RDMA networking; write automation with Python, Ansible, and Terraform; run health checks, benchmarks, and documentation; collaborate with cross‑functional teams to meet technical requirements.

Qualifications

• Strong hands-on Kubernetes administration and architecture.
• Bare Metal GPU cluster deployment and management experience.
• Kubernetes networking, storage, and security.
• Infrastructure automation (Ansible, Terraform, etc.).
• Strong Linux systems expertise and working knowledge of the NVIDIA GPU stack (CUDA, drivers) and RDMA.
• 2-4 years in systems/infrastructure engineering with Linux fundamentals.

Responsibilities

• Provision and configure GPU nodes: OS, drivers, CUDA, GPU Operator, and RDMA connectivity.
• Write and maintain automation using Python, Ansible, and Terraform.
• Support cluster deployments, run health checks and benchmarks, and document results.
• Monitor GPU infrastructure, respond to alerts, and resolve or elevate issues.
• Maintain runbooks and operational documentation.
• Collaborate with cross-functional engineering teams to deliver software solutions.

Skills

Kubernetes

Bare metal GPU

Networking

Storage

Security

Automation

Linux

CUDA

NVIDIA drivers

2-4 years experience

Tools

Ansible

Terraform

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 bare metal, on infrastructure built to keep GPUs available and models serving.

About the team

The Compute team owns Nava's GPU infrastructure end-to-end: from bare-metal provisioning of NVIDIA GPU systems through cluster software, RDMA integration, and production operation. We keep thousands of GPUs healthy and highly utilized so training and inference workloads run fast and reliably.

Responsibilities

• Provision and configure GPU nodes: OS, drivers, CUDA, GPU Operator, and RDMA connectivity.
• Write and maintain automation using Python, Ansible, and Terraform.
• Support cluster deployments, run health checks and benchmarks, and document results.
• Monitor GPU infrastructure, respond to alerts, and resolve or elevate issues.
• Maintain runbooks and operational documentation.
• Collaborate with cross-functional engineering teams to understand business and technical requirements and deliver software solutions.

Required qualifications

• Strong hands-on experience with Kubernetes administration and architecture.
• Bare Metal GPU cluster deployment and management experience.
• Kubernetes networking, storage, and security.
• Infrastructure automation (Ansible, Terraform, etc.).
• Strong Linux systems expertise and working knowledge of the NVIDIA GPU stack (CUDA, drivers, GPU Operator) and RDMA (RoCEv2/InfiniBand).
• 2-4 years in systems/infrastructure engineering with strong Linux fundamentals and a learning-oriented mindset.

Preferred qualifications

• NVIDIA GPU ecosystem exposure.
• AI/ML infrastructure experience.
• Monitoring tools such as Prometheus, Grafana, Dynatrace, Datadog, or Zabbix.
• Hybrid cloud and datacenter infrastructure experience.
• NCCL and NVLink tuning, Slurm, and distributed training/inference workloads.

Technology environment

NVIDIA GPU systems (DGX/HGX-class, Blackwell), NVLink, CUDA, GPU Operator; RoCEv2 / InfiniBand RDMA fabrics; Kubernetes (networking, storage, security), Slurm; Ansible, Terraform, Python; monitoring with Prometheus, Grafana, Dynatrace, Datadog, or Zabbix; Linux at scale.