Solution Architect - GPU & HPC

United States Digital Space LLC

Greater London

On-site

GBP 90,000 - 130,000

Full time

3 days ago

Be an early applicant

Application generator

Turn this role into an interview — a resume and cover letter built around what this employer wants.

Get past ATS filters

Benefits offered by this job

Competitive salary

Discretionary bonus

Wellbeing benefits

25 days holiday

Travel-required role

Job summary

NexGen Cloud, a leading AI cloud infrastructure company, seeks a Solutions Architect to own the technical sales cycle for GPU cloud workloads. You will translate workload requirements into architecture designs, collaborating with sales, data centre, and engineering teams to ensure feasible, deliverable solutions.

You will craft detailed designs, BoMs, and proposals, while shaping the product roadmap with customer feedback and competitive intelligence. Travel to customer sites is expected.

Qualifications

Proven HPC or AI software stack design at scale with workload profiling and scheduling.
Deep understanding of CUDA, cuDNN, NCCL and production AI workloads.
Experience optimising AI/HPC workloads across multi-GPU/multi-node systems.
Strong containerisation/orchestration knowledge (Docker/Kubernetes, NVIDIA Operator).
Exposure to full design-to-deployment lifecycle for GPU workloads.
Ability to document architecture and diagrams for technical/board audiences.
Comfort acting as technical authority in customer discussions.

Responsibilities

Own the technical sales cycle end-to-end as primary authority for GPU cloud designs.
Gather workload requirements and produce detailed solution designs with diagrams.
Collaborate with engineering teams to validate delivery feasibility before commitments.
Build reference architectures and templates for AI/ML workloads (training/inference/HPC).
Develop high-quality proposals, RFP responses, and SOWs with clear scoping and estimates.
Define BoMs for proposed solutions for procurement and margin review.
Provide feedback to leadership to influence product roadmap.

Skills

HPC workload design

GPU software environments

Multi-GPU tuning

Docker & Kubernetes

GPU workflow deployment

Technical documentation

Customer-facing authority

NCCL/MLPerf benchmarking

MLOps tools

InfiniBand networking

BoMs/Proposals/RFPs

Tools

SLURM

PBS

MPI

NCCL

PyTorch

JAX

DeepSpeed

Docker

Kubernetes

NVIDIA GPU Operator

Kubeflow

Airflow

Job description

Solutions Architect

Location: UK-based — Customer-Site Travel Required

ABOUT NEXGEN CLOUD

NexGen Cloud is the company behind Hyperstack, a full-stack AI cloud serving tens of thousands of customers from AI researchers to enterprises running the world’s most compute-intensive workloads. We deliver on-demand and private GPU infrastructure to teams who treat performance as a requirement, not a feature.

We’re a tight-knit, fast-moving team working at the cutting edge of AI cloud infrastructure. We practise what we preach, equipping our people with AI at every level so we can solve harder problems, ship faster, and keep raising the bar for what enterprise GPU infrastructure looks like.

THE ROLE: Solutions Architect

This role exists because our pipeline of technically complex, high-value GPU cloud opportunities is growing — and winning them requires more than a great sales team. The Solutions Architect sits at the intersection of sales, infrastructure, and the customer, translating complex workload requirements into technically sound, commercially viable solutions on the Hyperstack platform.

You’ll be the primary technical authority through the sales cycle: engaging directly with prospective and existing customers, producing detailed solution designs, and ensuring that what is proposed can actually be delivered to the standard committed. This is a role for someone who is equally comfortable in a customer meeting and a technical design review — someone who earns trust through depth, not slides.

WHAT YOU’LL BE DOING

Rather than a long checklist, here’s what success in this role looks like:

Own the technical sales cycle end-to-end — from initial customer brief through architecture design, proposal, and delivery handover — acting as the primary technical authority for GPU cloud solution design.

Engage directly with prospective and existing customers to understand workload requirements, technical constraints, and commercial objectives, producing detailed solution designs including architecture diagrams, network topology, storage configurations, and GPU resource allocation models.

Collaborate closely with Pre-Sales Engineering, Data Centre, Infrastructure, and Network Engineering teams to validate delivery feasibility before commitments are made to customers.

Build and maintain a library of reference architectures and solution templates spanning AI/ML training, inference, HPC, and rendering workloads — accelerating the sales cycle and improving proposal consistency across the team.

Develop high-quality technical proposals, RFP responses, and statements of work, supporting commercial discussions with clear scoping, realistic estimates, and honest risk assessments.

Define and maintain comprehensive Bills of Materials (BoMs) for all proposed solutions, ensuring accuracy for procurement, provisioning, and margin review.

Feed back recurring customer requirements and competitive intelligence to engineering leadership, directly influencing the Hyperstack product roadmap.

ABOUT YOU

We’re more interested in how you think and work than in a perfect CV. You’ll likely come from an HPC, AI infrastructure, or high-performance cloud background and bring a combination of the following:

Essential

Proven experience in HPC or AI software stack design and delivery at scale — including workload profiling, scheduler configuration (SLURM, PBS, or equivalent), MPI/NCCL tuning, and distributed training frameworks such as PyTorch, JAX, or DeepSpeed.

Deep understanding of GPU software environments: CUDA, cuDNN, NCCL, driver stacks, and the tooling required to run large-scale AI training and inference workloads reliably in production.

Hands-on experience optimising AI and HPC workloads across multi-GPU and multi-node configurations — including profiling, bottleneck identification, and performance tuning at both the application and infrastructure layer.

Strong working knowledge of containerisation and orchestration in HPC/AI contexts: Docker, Kubernetes, NVIDIA GPU Operator, and container-native workload management.

Background in an OEM, hyperscaler, neo-cloud, or enterprise/research HPC environment, with demonstrable exposure to the full design-to-deployment lifecycle for GPU-accelerated workloads.

Ability to produce clear, professional technical documentation and architecture diagrams suitable for both engineering and board-level audiences.

Confident engaging with customers, vendors, and internal engineering teams as a technical authority — able to translate complex software and performance trade-offs into clear, actionable decisions.

Nice to Have

Experience with large-scale cluster performance benchmarking — NCCL tests, MLPerf, or equivalent — and familiarity with what good looks like across different GPU generations and topologies.

Exposure to MLOps tooling and AI platform layers: experiment tracking (MLflow, W&B), model serving frameworks (Triton, vLLM), and pipeline orchestration (Kubeflow, Airflow).

Familiarity with InfiniBand and high-performance networking as it relates to distributed training performance — sufficient to engage credibly with network and infrastructure teams on topology and tuning decisions.