Project Context: At Limagrain, a business critical, internally developed app, had only one environment, which was manually built. I was in charge of building all environments, with automation.

• Tech stack : Terraform, Helm, kubectl, Docker, Azure AKS / App Gateway Ingress Controller (AGIC), Azure Functions, Azure CosmoDB, Azure EventHub, Azure Devops (Pipelines), Azure Managed identities / Service Principals, Azure networking - vnet, subnets, private endpoints, Artifactory, shell scripting (sh, bash).

I was first assigned the task of auditing the whole project, both in terms of infrastructure, and CI/CD, with an executive summary as deliverable. My two main recommendations were :

• to automate “everything”: both the infrastructure management with terraform, and the dev CI/CD
• to spin up isolated environments to promote releases from dev, to staging, UAT, up to the production.

Duties and achievements:

• I was in charge of implementing my recommendations, working with the Dev Team (5 developers), on one end, and the core infrastructure teams, on the other hand.
• I was invited to architecture labs, to give a talk on terraform and infrastructure as code concepts. My talk was very appreciated, so that other departments requested to have knowledge sharing workshops with me.
• I was required to attend the validation gates meetings: the process to allow the app to go to production. My role was to answer the question about infrastructure and automation design choices details.
• Terraform was not used at all in the project, and I convinced the management, and the core infrastructure teams, to start using terraform for infrastructure management, especially in production.
• In less than 4 months, I brought up with terraform, 3 new isolated environments: DEV, UAT and production. I also found and solved a couple of very complex issues integrating Azure cloud components and Microsoft Dynamics.
• The terraform achievements, my evangelist work, added to the support of the project managers, resulted in a real victory: In January 2024, several dedicated Azure subscriptions were granted to the Dev Team department, where initially a single Azure subscription had to be shared by all departments!

Project Context: At Michelin, Data Engineering / AI Project

• Tech stack : Terraform, Kubernetes, Helm, Docker, Azure, Databricks, dbx, Snowflake, dbt, azcopy, Gitlab, Artifactory, shell scripting, Python, TypeScript, Astro, Hugo.

In charge of managing all Azure resources with terraform (Azure Storage Accounts, vnet/sbnets, VMs, managed identities, service principals, etc…) Re-designing and implementing, with 2 Data engineer SMEs, the whole Data Engineering platform. I was working on the CI/CD pipelines scope. The new design involved a migration from an {Azure Data Factory / ADF Pipelines} stack, to a much more advanced, Python/PySpark-based, {Databricks + Snowflake} Stack.

The new data engineering platform processed data by executing Python code, and therefore needed a fully fledged Python CI/CD.

Duties and achievements:

• Managing of Azure resources with terraform: Databricks workspaces, Azure Account Storage, networks, Azure Key Vaults, Managed Identities, R-Studio VMs, MongoDB clusters, etc…
• Modularization of terraform code, management of terraform state: terraform modules, terraform private registry, terraform state into Azure Account storage container
• Design of CI/CD Modules, CI/CD runners Docker images (Python code)
• Design of CI/CD Pipelines: Git workflows, Gitlab release automation with gitlab-cli, python packages deployment to databricks (databricks scheduled jobs), snowflake (with liquibase).
• Design of Documentation: astro/hugo documentation repo with automation based on Gitlab CI/CD pipelines.

Context: Gravitee is an Open Source API Gateway

• Tech stack : Pulumi, Terraform, Kubernetes, Helm, Docker, K3D, Azure AKS, AWS EKS, FluxCD. NodeJS, TypeScript, Python, shell scripting, Golang.

I completely re-designed I re-implemented the whole Gravitee CI CD, migrating from Jenkins to Circle CI on the way. I today am a top expert about Circle CI and git based CI/CD. I am extremely proud that I had the opportunity to work with the https://gravitee.io Team, a World Class Team of Software Engineers, with which I deeply learned about many topics. I also there learned a lot about what cybernetics is in the real cloud industry world. In particular, it is there that I started working on Kubernetes Operators.

Duties and achievements:

• Design, implement, and migrate of the whole of the Gravitee CI/CD System. I designed and implemented a very unique CI/CD component, using NodeJS, TypeScript, RxJS, to orchestrate Circle CI Pipeline executions, for the Gravitee Release process (dozens of git repositories): bringing scalability, and removing Single Points of Failures.
• Automation of all standard operations of Gravitee, for hosted offers (deployment, backup, restore, upgrades, downgrades, etc).
• Design of saas offer fo the Gravitee API Gateway,
• Design of best architecture practices for Gravitee API Gateway users.
• CircleCI Evangelist for the Dev Engineers
• Training junior devops engineers
• Level 3 Support for Gravitee Customers, example : design and provisioning of a VPN Gateway between Azure and Google GCP

Project Context: Cresh offer is a SAAS Offer, an innovative b2b payment solution.

• Tech stack : Pulumi, Helm, Docker, Kubernetes, AWS EKS, Gravitee API Gateway. NodeJS, TypeScript, bash / sh shell.

At https://cresh.eu, a small french Startup, I worked on designing and implementing the entire Cloud Infrastructure, 100 % Kubernetes based. AWS EKS, Helm deployment of dev’s work, Gravitee API Gateway. https://gravitee.io founders heard of me because of the work I did there at cresh. Small Startup when I helped them, I was at the time the only devops engineer, and massively used Pulumi, Packer and Terraform. Having major architecture and technical points of disagreement with the CTO, we together agreed to end our collaboration.

Duties and achievements:

• Cloud Engineer, full infrastructure design
• Kubernetes SME
• Main Kubernetes Cluster Terraformation Automation.

Project Context: Edge Datacenter deployments for an ISP/Phone Operator.

• Tech stack : Ansible, Docker, Gitlab EE, Jenkins Pipelines, LVM, Cisco NSO, Cisco KUBAM for PXEless boot, SSH Bastions, Python .

Devops contractor at SFR (French ISP and phone operator), Cisco NSO project Team (SDN). In charge of datacenter deployment automation and operations in the context of the new 5G network deployment.

Duties and achievements:

• I worked on designing and deploying a full devops stack, for developers to git push their Python source code, ending with their code deployment to production Cisco NSO devices. I even had, on the SFR request, to build from source docker registry it self : SFR needed its CI/CD platform to be “fully built from source” (docker, registry, etc.), and that, for security constraints.
• I was a CI/CD evangelist on this Cisco NSO project (networks engineers do not have the same culture as Dev Engineers, to say the least). This included introducing and training network engineers on git, and why Gitlab and Github are our everyday Facebook (I used to say, Facebook is our today’s real office - and by “Facebook” I meant Github/Gitlab-like platforms).
• I designed git workflows, trained on how to operate Gitlab EE as an infrastructure on premise : software engineers at SFR, were in charge of operating their own Gitlab EE.
• I took a Cisco NSO training session at Cisco, on how to automate bare metal provisioning of Cisco NSO Clusters, in « Issy les Moulineaux » a French city close to Paris.
• With those duties, I worked with more than 100 engineers, coming from 3 different teams :
  • pure Infrastructure team (those who own the OpenStack),
  • software development teams,
  • network engineers teams

Context: Infrastructure Management Consulting services.

• Tech stack : Terraform, Ansible, Jenkins Pipline, Docker (docker-compose) , NGINX, Gravitee.io (API Gateway), Keycloak, Kubernetes, ELK.

Employee at Bosstek, a french consulting SMB, specialized into infrastructure management, that was my first experience in a « Pure Infra » company.

I took part into 3 projects with Bosstek, for 3 different French Customers. This experience was very interesting, for I worked there with people who never ever worked in software development. I would say they were from the « old style » infrastructure world, and that is why i had a lot to learn with them.

Duties and achievements:

• I worked in the core team of projects for huge companies : SOPRA Steria, ENEDIS, and Carrrefour One Cloud, as Cloud Engineer.
• I have been Dev Engineers Team Leader At SOPRA STERIA, where I worked under the authority of David Maurange.
• I also conducted two internal studies about in that company, where i was the only Kubernetes SME :
  • Production grade Kubernetes Cluster Ingress Controller best practices.
  • A proposed solution of system monitoring with Rsyslog/Elastic Stack, (with a focus on CA servers restarts) : purpose was to have the company certified as a CA Authority delivering SSL/TLS certificates.
• Working in the Carrefour One Cloud (a project in collaboration with Google), I there discovered for the first time the concept of Headless CMS, with https://strapi.io , and it is the first time I worked on a production grade use of the Gravitee API Gateway.

Context: R&D fault detection in fiber optics large networks (OTDUs).

• Tech stack : Ansible, PXE boot, PXE boot kickstart installations, Docker, Virtual Box, KVM, Free IPA Server, SSH / SCP/ WinSCP, with a lot of bin/bashing.
• CI stack : Gitlab, Artifactory, MAVEN3 Git, Eclipse JBoss Tools, Jenkins, Checkstyle, JMeter, JUnit, DbUnit, Jasmine + Karma, Wireshark & wireshark dissectors
• Dev stack : Java, JCA IronJacamar, JPA, Hibernate, WILDFLY, EJB3 Stateless & MDBs’, JAX-RS & Resteasy, Apache Shiro, Leaflet, Angular, Linux Pluggable Authentication Mechanism, mariadb, postgresql/postgis, .Net Windows Service Application component, Visual Studio.

Java Dev Enginer at Vierling-Communication, an R&D company of the « Laudren Group » . R&D main goal was to design and develop a cutting edge network monitoring system.

What was thrilling and extremely interesting in this experience, is that the system did not monitor software, but hardware, and in particular, the system goal mainly was to monitor Fiber Optics and Copper cables failures !

The developed Monitoring system consisted in both hardware and software : pure Linux OS, java and C language.

Duties and achievements:

• I worked on the software part which was the « brain » of the monitoring system : the server that collects all data from distributed systems, analyze them, and manages alerts, using JCA Components in a Jee JBoss EAP app.
• I introduced there bare metal provisioning : PXE- booted to deploy the whole stack, and i automated all that to be able to run reliable testing. That was my answer to « it is not possible to fully automate our solution deployment ».
• I introduced and convinced of the necessity of Acceptance Tests.
• I developed an authentication module implemented using Jee Listeners and Spring modularization
• I brought in Docker-ized deployments, the internal use of Pipelines using Gitlab on premises, the Java deployment happening in JBoss EAP (Angular 8/9 + bootstrap Web UI).
• Geolocation of hardware failures, so GIS matters, I introduced Leaflet on the Web UI This experience was a turning point for me : That is were I worked months with whom I consider my master as far as *NIX and what OS actually are, a « Linux Wizard » , Eric Lévenez, a man I will always remember.

Context: IT Dept. in a contruction industry company

• Operational : Gantt charts, Specific Terms and Conditions documents in building industry (CCTP in French) / Response to tenders (« Réponses à appel d’offres » in French) / supply chain management on a constuction site.
• Devops stack : Chef.io, Virtual Box SOAP API, KVM Virtual Manager, GNU/Linux (CentOS/Ubuntu server)
• CI stack : Artifactory, MAVEN3 (M2E), Git, Eclipse (eclipse bpmn modeler), Jenkins, Checkstyle, JMeter, JUnit, Mockito, DbUnit.
• Dev stack : Spring, JPA, Activiti BPMN engine, petals ESB / JBI java standard, tomcat 7, JAX-WS/ SOAP/ WSDL, JAX-RS, Spring WS, JSF2, Prime Faces, jQuery v2.x.x, JavaMail, Apache James server, Mailets.
• Misc. : FM 200 Fire protection systems… ?

SPVC was an SMB of the contruction industry, specialized into fluids.

I designed the first real Information System, conducted and took part in developments of software components required to deliver that first I.S.

I thereafter structured the newly created IT department to be fit for production management. All in all, this was among my most craziest professional experiences, that made me learn a lot about who i am in the IT industry.

Also my first experience as a pure Software guy, putting his hands unawarely into infrastructure.

• The Work began with the design and implementation of a simple utility software, whose goal was to normalize Business Processes, especially for the commercial and build sites managers everyday tasks.
• I then made it production « almost » ready, by adding to it the seed of a DRP. The implemented services were relation to construction project management : Gantt Charts, Tasks Management, reporting.
• A second generation, redesigned IS turned it into a B.A.M. system (Business Activity Monitoring) using :
  • The Acitviti BPMN engine.
  • A JBI ESB coupled to the BPMN engine, SOA architecture (all BPMN execution were invoked through the ESB, using JAX-WS / JAX-RS).
  • 15 users in January 2014, up to 30 end of 2014, then 40 end of 2015.
  • As I first talked of the critical importance of a DRP, i was certainly not realizing how deeply I would dive into infrastructure management…
• I built month weeks after weeks a small team consisteing of 3 to 5 developers and ops, I was the software architect with 2 developers, and manager for the 2 / 3 ops engineers (system enginrer, security engineer, network engineer).
• I was personnaly working on the CI system a classical recipe of Git, Maven, Jenkins (Junit, DbUnit, Mockito, etc…), Artifactory. All classics of that prehistorc times.
• One unthinkable thing happened there, where I learned a lot about real infratructure , I conducted and designed a full server room: SPVC used on premises servers, the IS being used by employees from their tablets mainly (MTT french trademark Android tablet). A crazy experience where i learned what a datacenter really is, e.g. what material are used on the floor (epoxy) of datacenters, Server Hardware Benchmark, network backbone, DRP with FM200 fire systems, air cooling, electricity etc…

Context: IT Dept. of one of the world’s biggest HR industry player.

• CI stack : Maven, Tortoise SVN, JIRA, Jenkins, Nexus Sonatype
• Dev stack : Tomcat 6, WebLogic 9, struts 1, Taglibs, Spring, hibernate, JUnit, JMX, JAXP, Log4J, Fractal Julia, Javamail.

At the heart of Michael Page’s IT Department (EMEA), I took part in the full revamping of Michael Page’s Information system, along with 10 à 15 Java Dev Engineers, 2/3 IT Architects, QA and 10 Business Analysts.

My personal scope was focused on the « backend » part of all software responsible for Candidates Processes Management, including the Candidates Assessment. This was my first professional experience into a major large scale IT Department of an International Corporation.

Duties and achievements:

• In collaboration with IT Architects, and Business Analysts, I designed and implemented the “candidate assessment” components. That made me an SME on the Candidate Assessment functional scope.
• Classic Java development, involving CI, using Maven, JUnit, on premise Nexus Sonatype repo. Some Java dependencies had to be « mavenized », some others already were.
• I took part in a huge Security Operation, because of a discovered major intrusion attack in Michael Page Information System, a long memorable pizza week-end…
• I started there being involved into topics of CI/CD and it the Infrastructure it relies on.