About the role
Job Description
About the Role & Mission
AI infrastructure is undergoing a generational shift. Disaggregated, raw GPU hardware must be transformed into multi-tenant, production-ready, and sovereign AI clouds - without locking enterprises into a single hyperscaler or hardware vendor. That is where k0rdent AI by Mirantis comes in.
k0rdent AI serves as the declarative "super control plane" that manages complex, heterogeneous GPU infrastructure from Metal-to-Model™. We are hiring a Head of Solutions Architecture to own the technical win, build our pre-sales organization from the ground up, and lead the architecture strategy on high-value, long-cycle compute deals.
This is not a demo-jockey role. You will be an executive technical partner to NeoClouds, AI-native startups, enterprise AI teams, research labs, and sovereign buyers. You will lead high-stakes architecture debates on interconnect topologies, build TCO models that challenge hyperscaler economics, construct reusable pre-sales machinery, and shape our core product roadmap directly from field intelligence.
What You’ll Own
* Build & Scale the Pre-Sales Engine (40%) * Hire, coach, and retain a world-class team of Solutions Architects and Field Engineers; establish the global pre-sales operating model as headcount and deal volume scale. * Build reusable, scalable field assets: discovery frameworks, reference architectures, TCO/benchmark models, POV playbooks, and RFP response libraries. * Set a high technical quality bar across the organization; run enablement so every SA can speak credibly to GPU architecture, high-speed fabrics, and container orchestration.
* Own the Technical Win in Strategic Deals (40%) * Partner with Account Executives as the technical executive on strategic enterprise opportunities with long sales cycles (6–18+ months) and multi-million-dollar ACV/TCV. * Execute structured qualification (MEDDPICC or equivalent) to map decision criteria, surface economic buyers, and build win plans around technical champions. * Architect end-to-end solutions across compute, networking, storage, and orchestration; deliver sizing, capacity plans, and TCO comparisons vs. public clouds and self-built alternatives. * Design and drive POVs/POCs: define strict success criteria upfront, run performance benchmarks, and convert results into commercial momentum.
* Serve as the Technical Voice of the Customer (20%) * Feed structured product, capacity, and feature requirements directly back to product management, engineering, and supply planning teams. * Partner closely with the NVIDIA field ecosystem (Cloud Partner program, reference architectures, joint pursuits) to win strategic accounts. * Influence roadmap prioritization, packaging, and GTM strategy based on real-world field feedback.
By joining us, you will lead the technical field engine behind k0rdent AI, shaping how the world's most innovative companies build and operate GPU-powered infrastructure at scale.
Qualifications
What We’re Looking For
* Pre-sales leadership experience at a NeoCloud, hyperscaler AI organization, or accelerated-hardware vendor
* Enterprise Deal Leadership: Proven track record leading technical sales in complex B2B environments, ideally with multi-year committed-capacity or reserved-capacity deal structures.
* Player-Coach Leadership: Track record of hiring, developing, and leading Solutions Architecture or Field Engineering teams—with the technical depth to lead the room on hard deals while scaling others to do the same.
* Hands-on AI/ML Infrastructure Experience: You have stood up or operated production ML workloads (distributed training and multi-node inference) and deeply understand performance bottlenecks across interconnect, memory bandwidth, I/O, and cluster scheduling.
* NVIDIA Stack & High-Speed Fabric Mastery: Fluency across Hopper and Blackwell architectures (H100/H200, GB200 NVL72, B200) and reference systems (DGX, HGX, MGX). Deep understanding of NVLink/NVSwitch domains, InfiniBand, Spectrum-X Ethernet, RDMA/RoCE, and DPUs.
Strongly Preferred:
* Hands-on experience with cloud-native orchestration for AI: Kubernetes (GPU operators, device plugins), KubeVirt, Metal3/Ironic bare-metal provisioning, and Slurm.
* Familiarity with high-throughput parallel/object storage systems for AI pipelines and data center physical constraints (power, cooling, rack density).
Additional Information
We are a Leader for Container Management in G2 (#2 after AWS)!