What You’ll Do

• Integrate compute, networking, storage, facilities, and platform layers into unified AI infrastructure architectures
• Serve as named Design Authority on major AI infrastructure programmes — owning the architectural integrity end-to-end
• Map AI infrastructure architectures to client business outcomes: cost-per-training-run, time-to-inference, utilisation efficiency
• Evaluate and recommend vendor-neutral component stacks: GPU compute (NVIDIA, AMD), networking, storage, and platform
• Design reference architectures for common AI infrastructure patterns: training clusters, inference farms, hybrid training/inference
• Coordinate between domain specialists (networking, storage, MEP, platform) to resolve cross-domain design conflicts
• Conduct architectural reviews and risk assessments at programme milestones
• Present architecture decisions and trade-offs to client CTO/VP-level stakeholders
• 
  

What We’re Looking For

Must-Haves

• 15+ years in infrastructure and solution architecture, spanning compute, networking, and storage
• 3-5+ years designing AI/ML or HPC infrastructure platforms at scale
• TOGAF, SABSA, or equivalent architecture framework experience
• Vendor-neutral evaluation capability across GPU compute, networking, storage, and platform vendors
• Client-facing design authority experience on programmes of significant scale
• Systems-level thinking — ability to understand how component interactions affect overall system performance
• 
  

Nice-to-Haves

• Experience with NVIDIA DGX/HGX SuperPOD reference architectures
• Understanding of AI platform software: Kubernetes, Slurm, Base Command Manager
• Knowledge of TCO modelling for AI infrastructure (build vs colo vs cloud)
• Experience with air-gapped or sovereign AI infrastructure deployments
•