vultr - Strategic Technical Account Manager GPU

• Performance Optimization & Scaling • Identify bottlenecks (network, storage, memory bandwidth). • Provide tuning recommendations for batch size, mixed precision, parallelization strategies, and checkpointing. • Help customers evaluate cost vs. performance tradeoffs (GPU mix, CPU pairing, instance types, cluster sizing). • Technical Relationship Ownership • Own the long-term technical strategy across assigned GPU/AI accounts, including hyperscalers, labs, and high-growth AI startups. • Host recurring technical review meetings, roadmap reviews, and optimization sessions. • Define scaling plans, future GPU reservation needs, and capacity forecasting. • Incident & Escalation Management • Partner with Support, SRE, Networking, NOC, and Product Management & Engineering to resolve high-urgency incidents. • Manage outage communications, corrective action plans, and postmortem reviews with customers. • Advocate for GPU reliability improvements and influence roadmap priorities. • Account Growth & Expansion • Identify opportunities for expanded clusters, high speed storage, or networking upgrades. • Support Sales with technical validation and architecture diagrams needed for expansion. • Customer Advocacy & Product Feedback • Provide structured feedback on existing and future GPU offerings, networking fabrics, storage platforms, and upcoming AI/ML platform features. • Partner with Product on early access programs (new GPUs, pipelines, orchestration, etc.). • 2–5+ years as an AI/ML Engineer, AI/ML Ops, Technical Account Manager, HPC Engineer, Sales/Solutions Engineer or relevant technical role. • Strong knowledge of GPU hardware architectures (NVIDIA/AMD), CUDA/ROCm, distributed training, and ML frameworks. • Experience with Linux tuning, networking (Infiniband, RoCE fabrics). • Experience with high-performance storage systems (DDN, NetApp, Vast, Weka, etc.). • Ability to communicate complex concepts clearly to both executives and engineering teams. • Prior experience supporting hyperscale, AI labs, or large cluster deployments is a plus. • Cloud Native Computing Foundation Certified Kubernetes Administrator (CKA) certification is a plus.