Staff GPU Networking Engineer - Kubernetes & DC

Mural Group; United States; Remote; USD 270,000 - 462,000; Full time

Benefits offered by this job

Equity component

Bonus folded into base

Job summary

Mural Group, a late-stage AI cloud company backed by NVIDIA, runs AI compute in data centers worldwide and is building a managed Kubernetes platform to turn fleets of bare-metal GPU compute into isolated, scalable clusters for AI training and inference.

The networking team owns end-to-end leaf-spine networks across sites, defining feature requests and automating lifecycles with Kubernetes-native tooling to ensure reliability and rapid recovery from partial failures in a multi-site environment.

Qualifications

• Production ownership of bare-metal leaf-spine data center networks with VLANs, VRFs and EVPN/VXLAN.
• Experience with Kubernetes controllers, operators or CRDs in Go, plus GitOps and multitenant infrastructure.
• Strong observability and network lifecycle automation.

Responsibilities

• Own networking across the platform from Kubernetes cluster networking down to the data center fabric and GPU interconnect.
• Define how networking features are requested, provisioned, changed, observed and removed.
• Automate tenant and node network lifecycles with Kubernetes-native tooling, including safe recovery from partial failures.

Skills

Bare-metal DC networking

Kubernetes networking/tools

Go - CRDs/Operators

GitOps

Multitenant infrastructure

EVPN/VXLAN

Tools

NVIDIA InfiniBand networking

NVIDIA Spectrum-X RoCE

InfiniBand RoCE networking

NVIDIA NetBox

Job description

Mural Group, a late-stage AI cloud company backed by NVIDIA, runs AI compute in data centers worldwide and is building a managed Kubernetes platform to turn fleets of bare-metal GPU compute into isolated, scalable clusters for AI training and inference.

The networking team owns end-to-end leaf-spine networks across sites, defining feature requests and automating lifecycles with Kubernetes-native tooling to ensure reliability and rapid recovery from partial failures in a multi-site environment.