Senior Solutions Architect, GPU Cloud GenAI – Infrastructure

NVIDIA

India, MumbaihybridPosted 2 months ago

Skill Required

CloudGenerative AIObservabilityCD pipelinesEngineeringKubernetesPrometheusnetworkingTerraformdesigningbuildingAnsiblePythondesignC++AzureCI/CDHelmAPIsRAGAWSGCPGoCICDFulltime

Key highlights

Competitive salaries and generous benefits package.
5+ years of hands‑on infrastructure/platform engineering experience with large‑scale GPU clusters (100+ nodes).
Bachelor's degree in Computer Science, Computer Engineering, or equivalent.
Role based in Mumbai, India.
Requires deep expertise in Kubernetes and Slurm for GPU workloads.
Opportunity to engage with C‑level executives and influence NVIDIA product roadmaps.

Role overview

NVIDIA is seeking an experienced Solutions Architect & Engineer (SAE) with deep expertise in large-scale GPU cluster infrastructure and generative AI enablement. As a pivotal member of the Infrastructure and Platform Engineering team, you will architect and build GPU cloud platforms (IaaS, PaaS, SaaS) that power the world’s most demanding AI workloads, working closely with enterprise customers to deploy and scale GPU infrastructure.

Responsibilities

Design and architect scalable IaaS, PaaS, and SaaS layers for large-scale GPU cluster environments (32+ HGX/DGX nodes), spanning compute, networking, and storage orchestration.
Build multi-tenant GPU cloud platforms with production-grade APIs, control planes, and platform services that abstract infrastructure complexity for end users and application teams.
Develop cluster orchestration pipelines using Kubernetes (GPU operators, device plugins, multi-tenancy) and Slurm, optimizing for performance, reliability, and resource efficiency at scale.
Define and implement best practices for GPU resource scheduling, isolation, quota management, and observability, ensuring secure multi-tenant isolation and compliance.
Advise customers on deploying and scaling generative AI workloads (LLMs, MLLMs, RAG pipelines) on your infrastructure platforms, translating AI requirements into infrastructure specifications.
Engage with C-level executives and infrastructure teams to understand requirements, deploy GPU clusters across on-premises and hybrid cloud environments, and drive platform adoption.
Collaborate with NVIDIA engineering teams to resolve deep infrastructure bugs, provide feedback on platform capabilities, and influence product roadmap decisions.
Partner with customer infrastructure teams to tune, scale, and optimize GPU clusters for cost efficiency, throughput, and AI workload performance.

Requirements

5+ years of hands‑on infrastructure or platform engineering experience, with demonstrated expertise designing and operating large‑scale GPU clusters (100+ nodes).
Deep expertise building IaaS, PaaS, and SaaS platform layers—architecting and developing infrastructure foundations, not consuming cloud services.
Proficiency in Kubernetes (GPU operator, device plugins, multi‑tenancy) and Slurm for HPC and AI workloads.
Hands‑on experience with infrastructure‑as‑code (Terraform, Helm, Ansible), CI/CD pipelines, and observability stacks (Prometheus, Grafana, DCGC).
Strong coding ability in Python and/or Go/C++, building platform tooling and automation from scratch.
Experience with cloud‑native networking (InfiniBand, RoCE, RDMA) and distributed storage solutions for GPU environments.
Excellent communication skills, credibly engaging both infrastructure engineers and C‑level stakeholders on complex technical and strategic topics.
Bachelor's degree in Computer Science, Computer Engineering, or equivalent experience.

Nice to have

Working knowledge of LLM, MLLM, and RAG frameworks and how they map to infrastructure requirements.
Hands‑on experience with model serving frameworks (Triton Inference Server, vLLM, TensorRT‑LLM) and inference optimization techniques.
Proven track record optimizing infrastructure for cost efficiency, throughput, and resource utilization in multi‑tenant production environments.
Deep understanding of distributed training concepts (data parallelism, model parallelism, pipeline parallelism) from an infrastructure perspective.
Experience deploying and managing GPU clusters in cloud environments (AWS, Azure, GCP) and on‑premises infrastructure at enterprise scale.

Benefits

Competitive salaries.
Generous benefits package.
Opportunity to work with forward‑thinking and hardworking colleagues in a rapidly growing engineering organization.
Exposure to C‑level executives and influence over NVIDIA product roadmap decisions.
Commitment to diversity and an inclusive, equal‑opportunity work environment.
Supportive policies that prohibit discrimination based on race, religion, gender, sexual orientation, age, veteran status, disability, and other protected characteristics.