Senior Infrastructure Architect – GPU Data Centers

About the role

QuTwo is a Finnish AI company building sovereign AI solutions for Europe, with a platform that runs workloads across classical systems, AI accelerators and emerging quantum hardware. It is looking for a Senior Infrastructure Architect who is a hands-on builder: someone who has actually racked servers, cabled the fabric and brought GPU clusters online. The role designs, deploys and runs the GPU infrastructure behind the AI platform.

Key responsibilities

• Plan and execute GPU data center build-outs end to end: rack and power design, high-density cooling, cabling and physical deployment of GPU servers and networking.
• Design and stand up high-performance compute fabrics (InfiniBand / RoCE), storage and out-of-band management for large GPU clusters.
• Own capacity, power and cooling planning together with colocation partners and site operators.
• Build the software layer: provisioning, firmware and driver lifecycle, monitoring and cluster scheduling.
• Bake in security from day one: physical access controls, network segmentation, hardened baselines, key and secrets management and tamper-evident logging.
• Ensure compliance with customer security and compliance requirements, including EU data residency and sovereignty, ISO 27001-aligned controls and customer security audits.
• Provide operational support and troubleshooting.

Required profile

• Proven, hands-on experience setting up GPU or HPC data center environments, having physically built and commissioned clusters.
• Strong grasp of data center fundamentals: power distribution, UPS, air and liquid cooling, rack density and site selection trade-offs.
• Fluent English; Finnish is a plus. Willingness to travel occasionally.
• Nice to have: background in high-security environments (finance, government, defense, critical infrastructure) or eligibility for security clearance, and experience negotiating with colocation providers, utilities or hardware vendors.

Required skills

• Networking: InfiniBand, Ethernet and secure network design.
• Linux and automation.
• Security frameworks such as ISO 27001, SOC 2 or Finland's Katakri criteria.
• Nice to have: NVIDIA DGX/HGX-class systems, liquid-cooled deployments or multi-megawatt sites.