About Osama M. Alqahtani

Senior IT Systems Engineer with deep specialisation in AI infrastructure design, deployment, and operations. Architected and delivered a production-grade, sovereign on-premises AI inference platform on Kubernetes, covering GPU cluster build-out, LLM serving, MLOps tooling, an LLM API gateway, and full-stack observability, and administered AI clusters across on-premises and Google Cloud Platform (GCP) environments. Broader expertise spans Kubernetes ingress and service mesh (Istio, NGINX, Traefik), enterprise virtualisation (VMware, Nutanix, Proxmox), compute (HPE, Dell), and storage (NetApp, Pure Storage). Holds industry certifications in NetApp, VMware, Oracle Cloud Infrastructure, and CompTIA Security+. Proficient in Python, Bash, and infrastructure automation with Ansible.

What I do

I build and run the infrastructure AI runs on, end to end: from procuring and racking NVIDIA GPUs to serving large language models to the people who use them. At GOSI I led the design and deployment of the organisation's first on-premises AI cluster, a sovereign inference platform on Kubernetes that serves models of up to 235 billion parameters entirely in-house, with no cloud APIs.

Around the models sits everything that makes them usable in production: an LLM API gateway with per-team keys, rate limits and cost attribution; MLOps tooling for experiments and model versions; observability from GPU utilisation to token throughput; and an air-gapped image pipeline, so the cluster runs with no internet connection at all.

The same work covers the rest of the data centre. I moved production workloads from VMware to Proxmox across two sites to cut licensing costs, and I run enterprise storage on NetApp, HPE 3PAR and Pure Storage, with disaster recovery at zero data loss.

Where I come from

I studied Computer Engineering at King Saud University. My graduation project, IIMA, was a smart glove that translates Saudi Sign Language, using motion and finger-bend sensors and a machine-learning model trained on a dataset our team built. Alongside the final stretch of the degree I was a research assistant at the Center of Smart Robotics Research.

I started in security, as a Security Specialist at Accenture, monitoring and investigating incidents in Splunk and ArcSight. In 2023 I moved into infrastructure at GOSI, first running virtualisation and storage, and from 2025 building its AI platform.

I also teach and speak. In 2026 I taught six sessions of Misk's AI Infrastructure track, rated 4.6 out of 5, and I give talks about what happens when AI gets big.

Work together

If you're building AI infrastructure and want a hand, the ways I can help are on Work with me. If you just want to talk, reach me on LinkedIn or write to omaalqahtani@gmail.com.

The short version

  • Senior IT Systems Engineer at GOSI, specialising in AI infrastructure
  • Built GOSI's first on-premises AI cluster: 8 GPUs, 7-node Kubernetes, models up to 235B parameters
  • Virtualisation and storage across VMware, Proxmox, NetApp, HPE and Pure Storage
  • Certified by NVIDIA, NetApp, VMware, Oracle, CompTIA and Cisco
  • BSc Computer Engineering, King Saud University
  • Taught Misk's AI Infrastructure track, rated 4.6/5; spoke at Tuwaiq Club and PyData Riyadh
  • Based in Riyadh, Saudi Arabia