At LinSoft Tunisia, I built and operated a single-node Red Hat OpenShift cluster for internal services and development workloads. I also created a self-healing platform that used Event-Driven Ansible and large language model root cause analysis to cut mean time to recovery from 90 minutes to under 2.
At LinSoft Tunisia, I engineered a runtime security subsystem using Falco and eBPF, reducing false positive security alerts by 60%. After a power outage and DNS failure, I restored all 34 OpenShift operators to healthy within 4 hours.
At TDS by Nomios, I integrated Trivy scanning into the build pipeline and helped reduce critical and high vulnerabilities by 80%. I also authored Kubernetes manifests and Helm charts and built monitoring dashboards that measured 99.5% availability over 3 months.
At Tunisie Télécom, I built a Python FinOps tool that analyzed pod-level CPU and memory usage and identified services over-provisioned by up to 3x. In academic projects at ESPRIT, I built an OpenStack private cloud with a hybrid Azure extension and a Jenkins CI/CD pipeline with security gates.
