Head of Engineering

Hundreds of latest generation NVIDIA GPUs, clusters to multiply, sites to open: Arkane Cloud is looking for its Head of Infrastructure to drive the growth of a French and sovereign GPU cloud.

Arkane Cloud

Arkane Cloud is a Lyon-based Cloud Service Provider that makes AI computing power accessible, affordable, and available in minutes. Thanks to our own GPU infrastructure, startups, researchers, and companies deploy their AI models without struggling with the technicalities:

• Up to 70% savings compared to an hyperscaler.

• A Plug & Play platform ready in less than 3 minutes.

Our ambition: to build a leading European player in AI infrastructure.

Why this role?

Specializing in AI-as-a-Service and GPU-as-a-Service, we have an infrastructure ranging from A5000 to NVIDIA HGX B300 (InfiniBand XDR, VAST Data storage). We are creating this position to support us in future deployments, then to multiply clusters and sites, with a team growing in parallel.

This is a builder's role: you are responsible for the end-to-end infrastructure, from ordering hardware to the GPU available to the client.

Your missions

• Design and deploy GPU clusters (HGX B300, GB300, Vera Rubin) according to NVIDIA reference architectures.

• Manage network and storage: InfiniBand XDR, RoCE v2, external connectivity, VAST Data.

• Select and negotiate data centers and suppliers: colocation, energy, liquid cooling, NVIDIA, OEMs, integrators.

• Maintain SLA: supervision, on-call duty, incidents, automation, ISO 27001 infrastructure component then SecNumCloud.

• Build and manage the infrastructure and DevOps team, and oversee TCO and revenue per GPU.

Your first 6 months

1st month: the HGX B300 cluster is in production, accepted, and compliant with the NVIDIA reference architecture.

3 months: on-call duty and incident process in place, 12-month capacity plan approved, key supplier contracts secured.

6 months: new cluster or site deployed on time and within budget, SLA met and measured, key recruitments made, ISO 27001 ready for audit on the infrastructure side.

Your experience

• 10 years or more in infrastructure, including at least 3 years leading a GPU infrastructure team.

• At least one cluster of several hundred NVIDIA GPUs (HGX) delivered in person

Your expertise

• High-performance networking: InfiniBand or RoCE v2 (rail-optimized, fat-tree, fabric management).

• Data center: colocation, electrical and thermal constraints, liquid cooling or over 120 kW per rack.

• Negotiation of hardware and colocation contracts.

• Production operations: SLA, on-call duty, incident management.

• Fluent professional English.

A significant plus:

• Experience at a cloud provider, a GPU neocloud, or an HPC player;

• Kubernetes and GPU orchestration (GPU Operator, MIG, Slurm);

• Infrastructure as code;

• High-performance storage;

• Inference stack (vLLM, SGLang);

Why join us

Choices to be made, not endured: you decide the architecture of latest-generation hardware.

A real leadership role: reporting to the CEO, with a direct impact on the company's trajectory and funding.

A function to build: processes, team, sites.

A French sovereign cloud: a European alternative to hyperscalers, with a lean team and quick decisions.

Package and benefits

• 100 to 120 k€ gross annual fixed salary

• BSPCE (Stock options)

• Flexible remote work, two days presence in Lyon every two weeks (reimbursed). Regular travel to data centers in Paris and to suppliers is expected.

• 100% health insurance, meal vouchers, RTT (Syntec Convention)

Arkane Cloud is committed to diversity: we consider all applications, regardless of gender, age, origin, or disability.

Find more English Speaking Jobs in France on Arbeitnow