Lead Solution Architect – GPUaaS

Astir IT Solutions, Inc.; Pennsylvania; Remote; USD 170,000 - 240,000; Full time

Job summary

Astir IT Solutions, Inc. seeks a Lead Solution Architect for GPUaaS to drive large enterprise engagements, design scalable GPU infrastructure for production AI inference, and establish reference architectures.

You will lead POCs, benchmarks, sizing, and performance optimization, while advising executives on capacity and cost. Strong pre-sales, leadership, and GPU expertise are required; remote work is supported.

Qualifications

• 15+ years in Enterprise Infrastructure, Cloud, or Solution Architecture.
• 5+ years designing, deploying, or managing GPU clusters (Mandatory).
• 1+ year selling GPU capacity, GPU Cloud, or AI infrastructure.
• 2+ years of production LLM inference deployment and operations using vLLM, SGLang, TensorRT-LLM, or NVIDIA NIM.
• Strong hands-on Python, Bash, Terraform, Ansible, and Helm.
• Expertise in NVIDIA GPUs (H100/H200, B200) and AMD Instinct.
• Strong experience with CUDA/ROCm, Linux, Docker, Kubernetes, and GPU Operator.
• Knowledge of GPU networking, storage, and multi-GPU architectures.
• Experience with GPU benchmarking, performance tuning, autoscaling, and multi-tenant environments.
• Understanding AI data center infra, high-density power, liquid cooling, and capacity planning.
• Experience with LLM optimization, model serving, token-based pricing, GPU sizing, and TCO.
• Strong customer-facing pre-sales, solution selling, exec comms, and tech leadership.
• Bachelor's degree in CS/Engineering or equivalent.

Responsibilities

• Lead GPUaaS solution architecture for large enterprise engagements.
• Design scalable GPU infrastructure and production AI inference solutions.
• Lead technical POCs, GPU benchmarks, sizing, and performance optimization.
• Advise executives on AI infrastructure strategy, GPU capacity, and costs.
• Develop reference architectures and technical standards.
• Collaborate with Sales, Product, NVIDIA, AMD, and partners.
• Mentor architects and lead complex solution design reviews.

Skills

Python

Bash

Terraform

Ansible

Helm

CUDA/ROCm

Linux

Docker

Kubernetes

GPU infra design

Pre-sales

Executive communication

Education

Bachelor's degree in CS/Engineering

Tools

NVIDIA GPUs

InfiniBand networking

Job description

Job Title: Lead Solution Architect – GPUaaS

Location: Remote

Minimum 15+ Years of Experience required.

Required Skills & Qualifications

• 15+ years of experience in Enterprise Infrastructure, Cloud, or Solution Architecture.
• 5+ years of hands‑on experience designing, deploying, or managing GPU clusters ( Mandatory ).
• 1+ year of experience selling GPU capacity, GPU Cloud, or AI Infrastructure, including multi-year commitments.
• 2+ years of production LLM inference deployment and operations using vLLM, SGLang, TensorRT-LLM, or NVIDIA NIM.
• Strong hands‑on Python, Bash, Terraform, Ansible, and Helm.
• Expertise in NVIDIA GPUs (H100/H200, B200) and AMD Instinct GPUs (MI300X/MI325X).
• Strong experience with CUDA/ROCm, Linux, Docker, Kubernetes, and GPU Operator.
• Knowledge of GPU networking (NVLink, InfiniBand, RoCE), storage, and multi‑GPU/multi‑node architectures.
• Experience with GPU benchmarking, performance tuning, autoscaling, and multi‑tenant GPU environments.
• Understanding of AI data center infrastructure, high‑density power, liquid cooling, and capacity planning.
• Experience with LLM optimization, model serving, token‑based pricing, GPU sizing, and TCO.
• Strong customer‑facing pre‑sales, solution selling, executive communication, and technical leadership experience.
• Bachelor's degree in Computer Science, Engineering, or equivalent experience.

Key Responsibilities

• Lead GPUaaS solution architecture and large enterprise customer engagements.
• Design scalable GPU infrastructure and production AI inference solutions.
• Lead technical POCs, GPU benchmarks, sizing, and performance optimization.
• Advise executives on AI infrastructure strategy, GPU capacity, and costs.
• Develop reference architectures and technical standards.
• Collaborate with Sales, Product, NVIDIA, AMD, and technology partners.
• Mentor architects and lead complex solution design reviews.

Certifications

• NVIDIA Certified Professional (NCP) – required or achievable within 12 months.
• Additional certifications: CKA, Red Hat, OpenStack, or Cloud Architecture.

Note: Candidates must have genuine hands‑on GPU cluster experience, production LLM inference experience, and GPU infrastructure selling experience. Pure cloud architects or AI/ML developers without GPU infrastructure expertise will not qualify.