Engineering Team Lead, Fleet Delivery

A datacenter is not a cloud until software makes it one. You will lead the team that turns arriving hardware into accepted clusters.

Cabling validation, burn-in, firmware, deployment validation, software stack install — all of it automated, none of it click-ops. Your KPI is time to acceptance: the hours between a rack being physically installed and it serving customer workloads. Across the industry that number is measured in weeks. You will drive it toward days, then hours, as we scale toward a gigawatt of compute across Europe.

This is a founding leadership role at the intersection of hardware and software. Early on it is deliberately hands-on: most of your time in the code and the architecture, some of it on datacenter floors, because the automation has to survive contact with the machines it runs on.

But the job is building and leading a team from day one — hiring your first engineers, coaching them, setting their growth paths, and running the rituals that keep quality high. As the team grows, the balance shifts toward people leadership, and you decide how fast. You shape this new team's charter and set the standard for how every future Mistral datacenter comes online.

• Build the delivery pipeline in Go: provisioning, configuration, validation, burn-in, network bring-up, software stack install.
• Build the team from scratch: make the first hires, onboard and coach them, own their performance and growth paths — and set the tooling standards and the playbook every future site reuses.
• Own time-to-acceptance and drive it down relentlessly — measure it, break it down, kill the slowest step first.
• Work shoulder-to-shoulder with Hardware Design and Fleet Health — reliability starts at delivery.
• Handle GPU fleets at scale: H100 and GB200-class nodes, InfiniBand and RoCE fabrics, and the failure modes they bring.
• Partner with Product Managers and the datacenter org to plan delivery waves against capacity commitments.