Aymen Belhaj

Senior DevOps Engineer | Azure Platform, Kubernetes & IaC

Senior DevOps Engineer with 4+ years building secure, automated cloud platforms across Azure, AWS and on-premises Kubernetes. At Linde, I manage Azure hub-and-spoke networking and hybrid DNS, delivered the migration of 15+ enterprise applications to AKS, and run Azure ML and GPU platforms with Bicep, Azure DevOps and ArgoCD GitOps, cutting development compute cost by 60%. Previously an OSS Engineer at Microsoft, resolving complex Azure networking and platform incidents for Premier and S&P 500 customers. Microsoft Certified: Azure Administrator Associate.

Lisbon, Portugal
LinkedIn
0
Languages
0
Skills
0
Certifications
Work Experience
Senior DevOps Engineer
Linde plcLisbon, Portugal
Sep 2023 — Present
  • Configured Azure hub-and-spoke networking and hybrid DNS resolution between on-premises datacentres and Azure, providing secure private connectivity for enterprise workloads
  • Delivered the migration of 15+ enterprise applications to three environment-isolated AKS clusters (development, test and production)
  • Standardised Azure ML infrastructure (workspaces, compute clusters, datastores) as reusable Bicep code built on Azure Verified Modules
  • Designed Azure DevOps multi-stage pipelines with gated promotion from development to test to production
  • Unified Helm charts with custom CRDs and per-environment overlays, deployed consistently across clusters with ArgoCD ApplicationSets
  • Reduced development compute cost by 60% through spot node pools with automatic on-demand fallback and scheduled eviction of non-critical workloads, backed by Azure budget alerts
  • Centralised secret management in Azure Key Vault using Secrets Store CSI Driver and Workload Identity, with automatic pod restarts on secret rotation
  • Enforced namespace isolation with Cilium network policies and transparent pod-to-pod mTLS
  • Designed and built a highly available on-premises Kubernetes platform for NVIDIA A100/H100 GPU workloads on Proxmox, with Kubespray, Cilium, a three-node etcd control plane and redundant HAProxy load balancers
  • Delivered a self-hosted LLM serving platform (vLLM serving Llama 3, Mixtral and Qwen3) with a RAG pipeline using TEI embedding and reranker models and Elasticsearch
  • Established a unified observability stack with Prometheus for cluster and GPU metrics, Loki and Tempo for log and trace correlation, and Grafana dashboards
Cloud Engineer (Part-time)
PixbetLisbon, Portugal
May 2022 — Aug 2023
  • Designed and deployed the platform on serverless AWS Lambda using Laravel Vapor, with SQS-based asynchronous background processing
  • Strengthened resilience and data protection with Multi-AZ RDS, encrypted S3 storage for payment certificates, and SES and SNS for email, SMS and two-factor authentication
  • Built a layered edge defence with Cloudflare WAF and custom OWASP rule sets, rate limiting, DDoS protection and admin lockdown zones, complemented by AWS Shield and Web ACLs
  • Implemented zero-trust source code access, restricting self-hosted GitLab to approved MDM-managed devices (Jamf, Apple Business Manager) with remote lock and wipe
  • Deployed and operated Endpoint Protector DLP to detect and block data exfiltration through email, USB and unapproved Git repositories
  • Reduced end-user response times with Cloudflare edge caching
OSS Engineer
MicrosoftLisbon, Portugal
Jul 2022 — Jun 2023
  • Handled complex production incidents for Premier and S&P 500 customers across App Service, Container Apps, Spring Apps, PostgreSQL and MySQL Flexible Server and Azure Storage, from rapid mitigation to formal root-cause analysis
  • Resolved enterprise networking issues involving VNet integration, VNet peering, Private Endpoints, Service Endpoints
  • Investigated failures with KQL queries across Azure service telemetry, memory dump analysis, CPU profiling and network trace analysis
  • Advised enterprise customers on automating deployments with GitHub Actions and Azure DevOps
  • Partnered with Customer Success Account Managers and engineering swarms on high-impact cases, and refreshed outdated technical documentation
Automation Developer (Internship)
PICadvancedAveiro, Portugal
Mar 2022 — Jul 2022
  • Trained a YOLOv5 computer vision model to detect XFP modules, fibre optic cables and test boards, integrated with a UR5e collaborative robot for autonomous handling
  • Built a Python socket-based control framework for real-time image evaluation and robot commands
  • Developed a monitoring web application with live task statistics and camera feed, hosted on an Azure VM
Skills
Azure AKSAzure MLAzure App ServiceAzure Key VaultAzure Workload Identity FederationAzure RBACEntra PIMAzure Cost ManagementVNet peeringPrivate EndpointsCilium network policiesCloudflare WAFBicepAzure DevOpsGitHub ActionsHelmArgoCDPrometheusGrafanaKustoKQLvLLMRAG pipelinesAWS LambdaAWS RDSRust ProgrammingPythonHub-and-spoke networkingHybrid DNSKubernetesCrossplaneClickhouseMDMDLP
Education
Bachelor of Science, Computer Science
University of Tunis El Manar
2016 — 2019
Engineering degree, Computer Systems, Networking and Infrastructure
University of Tunis El Manar
2019 — 2022
Master of Engineering, Computer Engineering
Universidade de Aveiro
Mar 2022 — Jun 2022
Certifications
Microsoft Certified: Azure Administrator Associate
Microsoft
Mar 2022
DevNet Associate
Cisco
Jan 2022
Microsoft Certified: Security, Compliance, and Identity Fundamentals
Microsoft
Mar 2020
Cybersecurity Essentials
Cisco Networking Academy
Sep 2021
Neural Networks and Deep Learning
Coursera
Feb 2022
Convolutional Neural Networks
Coursera
Mar 2022
Languages
English
Fluent
French
Fluent
Arabic
Native
Last updated 15 hours agoMade with freecv.org