1. Context
We build and run the GPU infrastructure behind MIRS™: our own proprietary, ultra-low latency data transfer technology that runs the heavy compute on our GPUs and delivers any app to any device. Unlike ordinary one-way data transfer, MIRS is two-way and interactive: anything you see or hear, you can act on in real time. One GPU serves many devices, and the experience feels local even though it runs in our data centers.
This role builds and operates the GPU backbone all of that depends on. This is an opportunity to own and shape the network and systems layer as we scale across regions, keeping the fleet fast, connected, and growing smoothly. The work is mostly office-based, with occasional trips to our data centers. You will have real ownership, room to shape how it is built, and a small, capable team around you to build it with. If you like greenfield work at real scale, this is a rare chance to help write how a platform like this runs.
2. Mission
Own and run the networking, systems, and infrastructure backbone of the entire GPU fleet across our data centers. You keep it fast, connected, secure, and always scaling, end to end.
3. Responsibilities
Run (ongoing operations)
1. Design and maintain the network environment: network topology and protocols, switch configuration, VLANs, IP schemes, dynamic routing (BGP), DNS/DHCP, and ACLs.
2. Operate the fleet's network-booted, diskless environment: PXE/iPXE/UEFI HTTP boot, network-root images, and overlay filesystems.
3. Build and maintain the OS images the fleet boots from: kernel and configuration, patching, and full image lifecycle.
4. Run daily operations: monitoring, performance and capacity tuning, incident response, and root-cause troubleshooting across network, OS, and hardware.
5. Apply security controls in line with ISO 27001: access control, change management, asset management, and hardening.
Build (projects)
6. Extend and harden the infrastructure as the fleet scales across data centers (high availability, capacity, performance).
7. Automate provisioning and repetitive operations.
8. Keep runbooks and network diagrams current so the whole team can operate the system.
4. Required Experience
You have designed and operated real network environments and large, distributed infrastructure across multiple data centers, and you are ready for a bigger canvas. Core areas:
Areas and Requirements
Networking
Data center networking experience: design and operation of network environments, switch configuration, VLANs, dynamic routing (BGP), DNS/DHCP, ACLs, bonding, and packet-level troubleshooting.
Operating systems
Proven experience maintaining and operating Linux servers at scale: OS lifecycle, patching, configuration, and troubleshooting.
Distributed systems
Hands-on with large distributed systems across multiple data centers, ideally multi-continental.
Storage
Deep storage knowledge: filesystems, IO and throughput tuning, capacity and reliability.
Hardware
Solid understanding of server hardware: GPUs, memory/RAM, and bare-metal components.
Security
Security practices in line with ISO 27001 and their operational implementation.
5. Beneficial Experience
It will be beneficial in this role if you also know:
• Programming or scripting.
• GPU clusters, HPC, or large bare-metal farms.
• Colocation and data-center operations.
6. Profile and Working Style
This is a broad, hands-on role with real ownership. The person who thrives here:
• Sees the whole stack: network layers and how the systems inside a single server type work together.
• Has a builder's mindset and is comfortable standing up new projects from the ground up.
• Keeps learning fast as the platform and environment evolve.
• Ranges across networking, operating systems, storage, and hardware, never boxed into a single track.
• Takes problems end to end and owns the outcome.
