ADEO

I worked on the Grafana observability platform for ADEO, supporting observability needs across the whole group.The platform handled monitoring and logging at large scale, with logs representing around 30 TB of ingestion per day. My role focused on operating and improving the observability stack so teams could reliably collect, query, visualize, and alert on production telemetry.My work included:<ul><li>Operating Grafana observability services at group scale</li><li>Supporting high-volume log ingestion, around 30 TB/day</li><li>Improving reliability and operability of the observability platform</li><li>Helping teams monitor production systems through dashboards, logs, metrics, and alerts</li><li>Working on platform-level observability used by many teams across the organization</li></ul>This experience demonstrates my ability to work on critical observability infrastructure at very large scale, where performance, reliability, cost awareness, and operational discipline matter.

I managed Grafana observability for ADEO at group scale, helping operate the monitoring and logging platform used across the organization. On the logging side alone, the platform ingested around 30 TB per day, requiring strong focus on reliability, scalability, cost control, and day-to-day operability.

Grafana

Kubernetes

Terraform

ADEO - Managing observability at a multinational scale

Fransys

<h1>Building and Operating a Production Kubernetes Platform</h1>Since 2022, I have designed, built, and operated the infrastructure platform behind Fransys.Fransys runs production customer workloads, including databases, on independent Kubernetes clusters managed by a central platform. The infrastructure is bare metal internally and cloud-based for customers, mainly on providers such as Hetzner, Scaleway, and Outscale.My role was focused on Platform Engineering: Kubernetes architecture, infrastructure automation, deployment workflows, observability, security, and reliability.The platform is based on Talos, Kubernetes, Terraform, Helm, GitOps, Harbor, GitHub/GitLab CI, Cilium, cert-manager, OpenBao, and the Grafana stack: Grafana, Loki, Mimir, Tempo, and OpenTelemetry.The deployment workflow is automated end to end: code is pushed to GitHub or GitLab, CI builds and publishes a container image to our Harbor, the Fransys platform receives the event, deploys the application with Helm to the right cluster, and monitors the result.The platform also handles cluster provisioning, configuration, deployments, rollbacks, telemetry, alerting, and secure-by-default standards such as zero-trust principles, network isolation, certificate automation, and centralized secrets management.Results:<ul><li>Initial customer deployment reduced from weeks to minutes or hours</li><li>Cluster provisioning in around 10 minutes</li><li>Monthly infrastructure maintenance reduced to almost nothing</li><li>Observed uptime above 99.95%</li><li>Stable production clusters with very few incidents</li></ul>This project demonstrates my ability to build and operate production Kubernetes platforms from scratch, automate infrastructure and deployments, improve developer experience, and maintain reliable cloud-native environments under real production constraints.

I built and operated the Kubernetes platform behind Fransys, turning production infrastructure deployment from a manual, weeks-long process into an automated workflow taking minutes to hours. The platform runs customer workloads on independent Talos clusters with GitOps deployments, centralized observability, secure defaults, and very low operational maintenance.

DevOps

Fransys Cloud - Multi-cluster production Kubernetes platform

Gridky

Challenge: Gridky was experiencing inconsistent deployment cycles and recurring production issues due to manual deployment processes and a lack of automated testing. This created delays and affected the overall reliability of the platform.Solution: I led a comprehensive overhaul of the deployment workflow by implementing a fully automated CI/CD pipeline using GitLab CI/CD. Key steps included:<ul><li>Establishing a Robust Test Suite: Introduced rigorous TDD practices to validate code changes before merging, reducing errors early in development.</li><li>Automation of Build, Test, and Deployment: Streamlined the entire deployment process using GitLab CI/CD, minimizing manual intervention and reducing human error risk.</li><li>Containerization: Adopted Kubernetes to ensure consistent environment management across development, staging, and production.</li><li>Team Collaboration: Instituted regular technical reviews and pair programming sessions to foster knowledge sharing and maintain high coding standards.</li></ul>Outcome: The new pipeline cut deployment times by 50% and significantly reduced production incidents. This not only enhanced the scalability and reliability of the platform but also empowered the team to work more efficiently and embrace continuous improvement. This experience at Gridky reinforced my expertise in DevOps and leadership, equipping me with the skills necessary to drive high-quality, automated software delivery in dynamic environments.

At Gridky, I served as Lead Developer & DevOps, managing a team of 10 to build and maintain robust, scalable APIs based on Laravel. I drove the adoption of Test-Driven Development (TDD) and implemented efficient CI/CD pipelines that streamlined deployments and enhanced code quality. My role also involved conducting technical reviews and mentoring team members to ensure high development standards and continuous process improvement.

Bastien Chalier

CTO as a Service | DevOps & Laravel Expert

About me

Work experience

CTO

DevOps | Developer

Lead développeur back-end | DevOps

Co-founder | CTO

DevOps

Portfolio

ADEO - Managing observability at a multinational scale

ADEO

Fransys Cloud - Multi-cluster production Kubernetes platform

Fransys

Gridky - Team Lead

Gridky

Talents

Recruiters

About