Hardware Deployment & Maintenance Engineer

Granite is building sovereign AI infrastructure for Europe — compute, energy, and a new generation of AI technology running in modular, liquid-cooled containers that it specifies, builds, and operates itself. The first containers went live in 2026. More are in build. Now Granite needs the person who keeps the hardware running.

This is a founding role. You will be the first person at Granite whose full-time job is the deployed fleet — which means you are not joining an existing maintenance function, you are building one. Within six months, you will have written the procedures, established the spares model, and set the standard that every container after the first one runs on. Within twelve months, you will have brought the fleet fully in-house.

You will report directly to the COO. The role is based in Copenhagen, Denmark, and is primarily remote with regular travel to sites in Denmark and internationally. This is a full-time position.

• →Build out and commission every new container: rack and stack servers, switches, and GPU nodes; run, route, and label copper and fibre; connect power; carry out burn-in and acceptance testing.
• →Keep the fleet running: diagnose and replace drives, fans, power supplies, motherboards, CPUs, optics, and cabling; run health checks and preventative maintenance across all sites.
• →Own spares and asset management end to end — inventory, tracking, and lead times across every site in the fleet.
• →Build the maintenance function from scratch within six months: procedures, as-built records per container, and the standard that every subsequent container runs on.
• →Take full operational ownership of the fleet in-house within twelve months.
• →Hold the technical conversation with hardware suppliers and installation contractors on Granite's behalf — on servers, networking, storage, and anything else the infrastructure depends on.

• At least five years maintaining hardware in live data centres, with hands-on exposure to GPU compute, high-density racks, and liquid cooling.
• Demonstrated experience installing and commissioning compute hardware on site, not just maintaining it once it is in place.
• Comfortable replacing whatever fails: drives, fans, power supplies, motherboards, CPUs, optics, and cabling.
• Runs copper and fibre properly — install, route, label, test, and trace faults.
• Has written procedures that others work from, and has owned spares management and lead time planning.
• Able and willing to travel regularly to sites in Denmark and internationally.
• Can hold a technical conversation with a hardware supplier on servers, networking, and storage.
• Lives in Denmark, or in the Nordics, Germany, or Poland with a straightforward route to Danish sites. Must have the right to work in the EU and hold a valid driving licence.

The following are useful but not required:

• Production depth in single-phase immersion, direct-to-chip, or rear-door cooling systems.
• Experience with RMA and vendor warranty processes.
• Basic Linux administration and remote hardware management.
• Scripting routine checks and diagnostics.
• Experience with containerised or prefabricated data centre deployments.
• Danish or another Nordic language.

Granite is building the infrastructure layer that European sovereign AI depends on — compute, energy, and AI technology running in modular, liquid-cooled containers that Granite specifies, builds, and operates itself. The first containers went live in 2026. The team is small, the work is concrete, and the infrastructure is real.

The hiring process for this role runs as follows: an intro call with Bifrost, an interview with the COO, a technical interview, a final conversation with the CEO, and then references and offer.