Maxence L.

Designed and operated a scalable, event-driven Kubernetes infrastructure, centralizing AI inference management across a global fleet.

Built elastic fleet management from the ground up, currently operating 500+ Kubernetes clusters via GitOps (ArgoCD) with automated contextual configuration, integrated security workflows, and pre-release validation.

Built Infrastructure as Code (IaC) with standardized conventions to enable rapid, consistent deployments across multiple regions.

Delivered a robust monitoring and observability stack for both internal and remote fleets, ensuring reliability and fast incident response.

Implemented a CI runner infrastructure, reducing build times by 30% and improving delivery velocity.

Optimized GCP infrastructure costs by leveraging spot instances and managing GPU resources (50+ T4/L4 GPUs per month).

Scaled workloads to handle 1M+ AI inferences per month.

Implemented a secure VPN solution with restricted, encrypted remote access for distributed teams.

Designed and implemented a robust CI/CD solution using GitLab CI and ArgoCD for streamlined software delivery.

Managed and maintained Kubernetes clusters on Orange Flexible Engine (Huawei Cloud) to ensure optimal performance and scalability.

Conducted internal security audits and established rigorous security protocols to safeguard system integrity.

Orchestrated the implementation of a comprehensive governance system, encompassing access control policies, change management

protocols, and efficient approval workflows.

Enforced the use of standardized Docker images, including Distroless, and facilitated seamless updates through tools like Renovate Bot.

Managed feature branch reviews via pull requests, fostering collaborative development and code quality.

Implemented a Finops strategy resulting in reduced machine usage and improved cost efficiency.

Crafted a one-year multi-cloud infrastructure roadmap, harmonizing product and technical considerations. Key focus areas included security

enhancements, centralized authentication, authorization, code review processes, scalability, internal tool optimization, and initiatives to

reduce developers' time to market, enhance usability, and expedite deployment.

Full Remote

Design of a Hybrid CI solution ( Buildkite ) Development of an internal Buildkite autoscaling solution based on Event Driven Managing Kubernetes clusters with an eco-system ( GKE ) Deployment of a CD solution with tenant ( Argo Workflows ) Design of Istio with canary upgrades Support teams on cross-functional topics based on DevOps culture