DevOps engineer in Tunis, Tunisia. I build cloud and GPU infrastructure for AI/ML teams and automate the operations around it.
- Platform engineering across AWS, Azure, GCP, OpenStack and VMware, with Terraform as the common thread and Argo CD for delivery.
- GPU infrastructure for AI/ML on NVIDIA and AMD: GPU-enabled Kubernetes nodes, driver and operator lifecycle, utilisation monitoring with DCGM.
- Current focus: cloud cost engineering (FinOps), migration automation, and shifting security left in CI/CD with Trivy, SonarQube and Infisical.
- Upstream, most of my work goes into llm-d, Kubernetes-native distributed inference, alongside fixes across the Kubernetes and cloud provider ecosystem.
Credentials are verifiable on LinkedIn.
 llm-d 12 merged |
 kubescape 10 merged |
 pulumi 3 merged |
 aws-controllers-k8s 2 merged |
 NVIDIA 1 merged |
 GoogleCloudPlatform 1 merged |
 FluidifyAI 1 merged |
 k0sproject 1 merged |
 Dokploy 1 merged |
 falcosecurity 1 merged |
 1devspace 1 merged |
Open pull requests in 23 more organisations:
- Reviewed 62 pull requests in llm-d/llm-d-router, loft-sh/vcluster, llm-d/llm-d and 18 more.
- Opened 38 issues in llm-d/llm-d-router, loft-sh/vcluster, finopsfoundation/focus_validator and 16 more.
All 72 projects, including 85 open pull requests
…and 32 more.
| Project | What it shows | Stack |
|---|---|---|
| azure-go-task-manager | Go microservices delivered to Kubernetes through Argo CD and Kustomize overlays | Go, Kubernetes, Argo CD, Kustomize, Azure |
| gcp-microservices-tf | Terraform that stands up GKE and hosts Google's Online Boutique demo end to end | Terraform, GKE, GCP |
| hybrid-cloud-monitoring | One Terraform setup provisioning AWS and Azure resources with a single Grafana view across both | Terraform, Grafana, AWS, Azure |