Director of Cloud Operations (US Neocloud)

About the role

Mirantis delivers and operates infrastructure for AI neoclouds: GPU-dense data centres, Kubernetes-based platforms (k0rdent) and the operational services that keep them running. You be part of a follow-the-sun leadership model and will own day-to-day operations for the US time zone.

You will be accountable for uptime, incident response, capacity and customer outcomes. You will lead a team of engineers and operators, and stay close enough to the technology to make sound calls during an incident.

Responsibilities

• Own operational delivery for US neocloud customers: availability, SLAs, incident and problem management, and change control.
• Lead and develop the US operations team, including engineers, NOC/SRE staff and site-level coordinators. Set hiring plans, on-call structure and performance standards.
• Run GPU cluster operations: health monitoring, firmware and driver lifecycle, fabric and storage performance, node replacement and capacity planning.
• Act as leadership escalation point for major incidents and lead the communication with customer executives.
• Work with customer data centre and engineering teams to define runbooks, handover processes and operating models.
• Work with Sales, Solutions Engineering and Product to scope operational services, support pricing and resolve delivery risk before contracts are signed.
• Define and report operational KPIs (uptime, MTTR, ticket ageing, GPU utilisation, SLA attainment) to Mirantis leadership and customers.
• Manage vendor relationships with hardware, networking and data centre providers.
• Improve operations through automation, standard tooling and a single operational view across sites.
• Coordinate with our European and APAC operations teams for follow-the-sun coverage and consistent practice.

Qualifications & Requirements

• 10+ years in infrastructure or cloud operations, including 5+ years leading technical teams and at least 2 years managing managers or multi-site teams.
• Experience managing teams operating AI/GPU infrastructure at scale (NVIDIA H100/H200/GB200-class or equivalent), including InfiniBand or RoCE fabrics, DCGM and scheduler/orchestration layers such as Kubernetes or Slurm.
• Strong conventional infrastructure background: Linux, networking, storage, virtualisation or bare-metal provisioning, power and cooling fundamentals.
• Production experience with Kubernetes and the surrounding tooling (observability, GitOps, infrastructure as code).
• Track record running a NOC or SRE function with defined SLAs, on-call rotations and formal incident management (ITIL or equivalent).
• Experience working directly with customers or tenants during incidents, including at executive level.
• Ability to read and challenge technical designs, and to explain operational trade-offs to non-technical stakeholders.

Preferred experience

• Experience at a neocloud, GPU cloud, hyperscaler, colocation or managed services provider.
• Familiarity with bare-metal lifecycle tooling (Metal3, MAAS, Ironic or similar) and multi-tenant GPU platforms.
• Experience building or scoping managed-service offerings, including SLAs and pricing inputs.
• Experience with Prometheus, Grafana, VictoriaMetrics, OpenTelemetry or comparable observability stacks.
• Certifications such as CKA/CKAD, ITIL 4 or relevant NVIDIA credentials.

Leadership expectations

• Sets clear priorities and holds the team to them.
• Hires well and develops people. Can build a team from a small base.
• Makes decisions under pressure and takes ownership of outcomes.
• Communicates plainly with engineers, customers and executives.
• Works across Customer Success, Sales, Engineering and Product without needing formal authority.