About the role
TensorX, a sovereign AI infrastructure platform in Dublin, seeks an Infrastructure Operations Engineer to build, patch and monitor our GPU-enabled server estate. You will work in a hands-on, contributor role supporting Dell PowerEdge hardware, iDRAC, BIOS and SAN/storage while collaborating with platform and inference engineers.
You will primarily be in our Dublin office, with on-call duties and occasional data centre visits.
Server build & configuration: Install, configure and update OS on servers, build to a standard. Patch management: Keep OS, BIOS, iDRAC and firmware baseline, run maintenance with rollback plan. Observability: Set up monitoring, logging and alerting for servers, storage and network. Error management & on call: Triage hardware/OS errors, fix or escalate, document cause. iDRAC & out-of-band management: Configure iDRAC, diagnose and recover servers remotely. Hardware faults: Diagnose failed disks, DIMMs, PSUs, fans, NICs and boards, swap parts. Network & firewall changes: Day-to-day changes on Dell switches, VLANs, port speeds. Storage: SAN and servers: LUNs, iSCSI, multipath, health and capacity. Racking & cabling: Rack, stack and cable servers to standard. Documentation: Keep asset register, rack diagrams and runbooks accurate. Documentation: Keep the asset register, rack diagrams, cabling records and runbooks accurate. 3+ years of professional experience as a server, systems or infrastructure engineer. Hands-on experience with Dell PowerEdge servers: iDRAC, Lifecycle Controller, BIOS setup, firmware updates and hardware diagnostics Linux administration on Ubuntu or Debian: OS installs, package updates, systemd, logs and basic shell scripting Experience running maintenance windows on production servers. Experience with monitoring and logging tools (Prometheus, Grafana, Loki, ELK or similar) Confidence making network changes at a switch command line: VLANs, trunks, LACP and port speeds Confidence making storage changes: SAN, iSCSI, RAID and multipath Racking and cabling to a professional standard, including fibre and optics A methodical approach to troubleshooting. You change one thing at a time and write down what you find Comfortable using AI-assisted development tools (e.g. Claude Code, Codex) as part of daily workflow A clear and concise communicator with engineers, datacentre staff and vendor support 25 days paid annual leave Hybrid Dublin office Datacentre visits Free inference tokens