Experience
Four years across three organisations, moving from cloud governance and IaC into GitLab platform ownership and GitOps delivery.
Roles
Senior Associate — Infrastructure, Platform Engineering
- Own GitOps delivery for backend services on Argo CD, standardising promotion across environments and removing manual per-environment configuration for client-facing platforms.
- Authored reusable Helm charts and an App-of-Apps architecture so product teams onboard services self-service — cutting config drift and platform-team ticket load.
- Built a DevSecOps CI pipeline with SAST, Gitleaks and Trivy as shared reusable stages, shifting source, secret and container scanning left across every consuming repo.
- Escalation point and SME for deployment failures, Helm templating and Argo CD sync errors across application and platform teams.
DevOps Engineer — GitLab CI/CD & Kubernetes Platform
- Administered and tuned the GitLab Runner fleet at the Linux OS level, resolving contention and resource bottlenecks to materially improve pipeline throughput across shared CI capacity.
- Planned and executed production Amazon EKS upgrades with minimal downtime — owning workload validation, rollback readiness and service-availability verification end to end.
- Implemented Istio service mesh with Argo CD canary deployments for progressive delivery, traffic shaping and rapid rollback on production microservices.
- Led RabbitMQ and Apache NiFi production upgrades with zero-to-minimal downtime, protecting business-critical integration services.
DevOps Specialist — Cloud Platform, IaC & Governance
- Established AWS Control Tower landing zones, Service Control Policies and IAM Identity Center SSO across a multi-account estate — owning access control and account structure, not just build automation.
- Built Terraform provisioning pipelines driven from GitLab CI/CD, automating repeatable infrastructure delivery across AWS and GCP.
- Delivered cloud optimisation work that cut spend by 30% while building secure AWS foundations for development and production.
- Developed reusable Ansible roles with dynamic inventory and Molecule test coverage; automated compliance checks saving ~25 audit hours per month.
Case studies
Problem → approach → outcome. All from documented production work, not side projects.
GitOps delivery with Argo CD and App-of-Apps Publicis Sapient
Problem
Backend services were promoted across environments manually, so every environment needed its own hand-maintained configuration. That produced drift between environments and a steady stream of tickets into the platform team.
Approach
Standardised promotion on Argo CD and authored reusable Helm charts behind an App-of-Apps structure, so a new service is onboarded by declaring it rather than by hand-configuring each environment.
Outcome
Product teams onboard services self-service. Configuration drift between environments is reduced and platform-team ticket load is lower.
GitLab Runner fleet tuning Deutsche Telekom Digital Labs
Problem
Shared CI capacity was contended. Developers waited on pipelines, and the bottleneck sat below the CI layer — in how runners were provisioned and how the host OS handled concurrent jobs.
Approach
Administered and tuned the GitLab Runner fleet at the Linux OS level, resolving resource contention and bottlenecks rather than simply adding more runners.
Outcome
Materially improved pipeline throughput and reduced developer wait time across shared CI capacity.
Production EKS upgrades and progressive delivery Deutsche Telekom Digital Labs
Problem
Production Kubernetes clusters needed version upgrades without taking down business-critical microservices, and rollouts had no safe intermediate step between deployed and not deployed.
Approach
Planned and executed EKS upgrades end to end — workload and dependency validation, rollback readiness, service-availability verification. Introduced Istio with Argo CD canary deployments so traffic could be shifted gradually and rolled back quickly.
Outcome
Upgrades completed with minimal downtime. Progressive delivery gave production rollouts a safe, reversible path.
Multi-account AWS governance OpsTree Solutions
Problem
A multi-account AWS estate needed consistent account structure, access control and guardrails — governance, not just build automation.
Approach
Established Control Tower landing zones, Service Control Policies and AWS SSO federation, and drove Terraform provisioning pipelines from GitLab CI/CD so infrastructure delivery was repeatable across AWS and GCP.
Outcome
Cloud spend reduced by 30%. Compliance automation saved approximately 25 manual audit hours per month.