Network Deployment Engineer; AI Factory Delivery Job Abu Dhabi UAE/Dubai,IT/Tech

Position: Network Deployment Engineer (AI Factory Delivery)
Full-time | Permanent | Start ASAP United Arab Emirates (UAE) – Remote UAE-based remote working About Sharon AI

Sharon AI is a neocloud, delivering trusted AI infrastructure organisations need to build, train and run AI support customers across the full AI lifecycle, from training through to inference and agentic AI, drawing on a strong ecosystem of technology and co-location partners to deliver capability where it's needed. As the first neocloud to join NVIDIA's AI Compute Program, we're growing quickly, scaling our AI Factory platform to meet rising demand for advanced compute.

The Role

As Sharon AI continues to grow, we're looking for a talented Network Deployment Engineer to join our UAE Delivery squad and play an important role in deploying and delivering the network layer of our AI Factories, reporting to our Regional Delivery Manager.

AI Factories are the high-performance GPU compute environments that underpin Sharon AI's AI cloud, serving customers and internal teams running large-scale training, fine-tuning and inference. In this role, the Network Deployment Engineer will take the network and GPU fabrics of each AI Factory from preparation through build, validation and platform deployment to customer acceptance, with a strong focus on NVIDIA SuperPOD reference architectures, RDMA and Infini Band.

The role is focused on delivery rather than day-to-day operations: you'll both deploy and test, and look for ways to make each deployment faster than the last. You'll work closely with Infrastructure and Platform Deployment Engineers, on-site Data Centre Deployment Engineers, the Platform and Infrastructure teams and Operations, with exposure to cutting-edge technology at the forefront of AI infrastructure.

What You'll Be Doing

Bringing switches under Netris using the standard configuration
Deploying and configuring Infini Band fabrics and leaf-spine (Clos) Ethernet topologies based on NVIDIA SuperPOD reference designs, including BGP and EVPN/VXLAN where used
Configuring and tuning RDMA networking (RoCE and Infini Band verbs) to meet agreed design and performance targets for distributed training
Implementing network segmentation, redundancy and fault isolation for multi-tenant environments, and working with platform engineers on network-aware scheduling in Kubernetes or Slurm
Directing on-site staff and the provider's remote hands to find and fix optics and cabling faults, and checking link health, optics and error counters across the fabric
Planning and executing network testing, including acceptance testing (ATP), integrated system testing (IST), fabric validation, RDMA bandwidth and latency tests, and NCCL benchmarking at increasing scale
Supporting fabric soak and cluster-level testing (L12), and working with GPU and switch vendors to fix defects before handover
Delivering repeatable deployments using infrastructure-as-code (Ansible, Terraform) and shared deployment scripts, timing each step and proposing changes that shorten the cycle
Feeding lessons learned back into standard playbooks and tooling, and to the Platform and Infrastructure teams
Integrating network monitoring and telemetry (latency, congestion, packet loss) so each AI Factory is visible to Operations from day one
Preparing handover documentation, runbooks and as-built records, supporting Operations during the hypercare window (up to one month), and reporting progress, risks and issues to the Regional Delivery Manager

What We're Looking For

4 to 8+ years' experience in network engineering, preferably in HPC or cloud environments
Hands-on experience with Infini Band and/or RDMA networking
Experience deploying and operating large-scale data centre network fabrics
Experience supporting GPU or AI/ML infrastructure at scale
Familiarity…