About this role
Compute Engineering builds and operates G-Research’s OpenStack private cloud and VMware vSphere estate. Spanning thousands of hypervisors and bare metal nodes across multiple regions, the OpenStack platform supports research, GPU and general compute workloads and is deployed using Kayobe and Kolla-Ansible. We’re looking for an experienced engineer with deep OpenStack expertise to support existing production environments and help build new ones. A key focus will be a major ongoing project to deliver virtual desktop infrastructure (VDI) and other interactive user workloads across virtual machines and Ironic-managed bare metal. Key responsibilities of the role include: Building and maintaining new and existing OpenStack environments using Kayobe and Kolla-Ansible Supporting production OpenStack regions across Nova, Neutron, Ironic, Cinder, Glance, Keystone, Designate and Octavia, acting as an escalation point for our Compute Operations team Extending Ironic to new hardware, including enrolment, inspection, cleaning, BIOS and firmware management and multi-tenant networking Delivering VDI and other workloads on VMs and bare metal, including image builds, UEFI and Secure Boot, vTPM and GPU-backed instances Tuning flavours, scheduling and placement for interactive, performance-sensitive workloads Building tooling and automation that reduces toil and makes routine and break-fix work safe and repeatable Improving platform monitoring and observability using Prometheus and Grafana Validating changes with Tempest, Rally and other automated testing tooling and rolling them out in stages Applying security hardening and timely vulnerability remediation
Requirements
The ideal candidate will have the following skills and experience: Extensive hands-on experience running OpenStack in production at scale, including building new environments and at least one major version upgrade Experience deploying with Kolla-Ansible or Kayobe Production experience with Ironic and bare metal provisioning Strong Linux skills, particularly Ubuntu, with a methodical approach to troubleshooting from userspace down to the kernel Solid experience with Ansible, Terraform, Git and Jenkins Practical experience using AI tools to build internal tooling, automate workflows, enable agentic operations and speed up day-to-day engineering work Good understanding of Neutron, its internals and how it maps to physical data centre networks Server hardware knowledge, including BMCs, Redfish, encryption at rest technologies Clear communication and experience working alongside operations teams Desirable experience: Running Windows guests on KVM or OpenStack, or deploying Windows to bare metal using Ironic, virtio drivers, cloudbase-init and sysprep Experience with VDI across OpenStack or other platforms Familiarity with GPU passthrough or vGPU using Nova Building tooling and pipelines to automate VMware operations, using Ansible or the vSphere APIs Knowledge of Ceph, including RBD and RGW Experience patching OpenStack services and maintaining downstream patches across releases Hands-on experience operating VMware vSphere in production, including ESXi, vCenter, HA/DRS and major version upgrades