Alexandre Marce

Contexte & Enjeux: Au sein de l’équipe AIPC (AIProduction Center) j’ai la charge d’une équipe de 11 personnes en charge du support sur les produits Data/IA déployés sur du Kubernetes. J’ai mis en place toute la pratique de support: • Rédaction de procédures • Rédaction de post Mortem • Training des équipes • Création de KPI • Mise en place des rituels d”quipe Dispositif: • Taille d’équipe : 40 personnes (Développeurs, DevOps, Architectes, Product, Sécurité) • Interlocuteurs clés : CTO, Tech lead, Product owners • Périmètre : Gestion d’équipe, Gestion des KPI, coordination des équipes, recrutement des • Criticité : Applications business critiques / forte volumétrie / exigences de haute disponibilité 24/7 (Serious) Objectifs : • Réduire le temps d’incident • Former les équipes aux différents produits • Produire des outils pour réduire les actions manuelles Principales responsabilités & réalisations : • Management d’équipes • Recrutement • Formalisation des bonnes pratiques d’incident • Formalisation du processus de lancement de projets Résultats / Impact : • Division par 4 du temps de réponse aux incidents • Montée de l’équipe de 5 à 11 • Développement de KPI objectifs pour suivi de l’équipe (MTTR, TTFR)

Contexte & Enjeux : Dans un contexte de migration, le client souhaitait héberger ses charges sur le cloud. Le choix du cloud étant incertain, la solution de Kubernetes a été retenue. Je suis intervenu pour aider les équipes à structurer les clusters, standardiser les méthodes de déploiement et de packaging au sein des équipes de développement et piloter le projet pilote de clusters sur le cloud AWS Dispositif : • Taille d’équipe : 40 personnes (Développeurs, DevOps, Architectes, Product, Sécurité) • Interlocuteurs clés : Directeur technique, Tech lead, Developpeurs • Périmètre : Architecture Cloud, CI/CD, Kubernetes, automatisation infrastructure, observabilité • Criticité : Applications business critiques / forte volumétrie / exigences de haute disponibilité Objectifs : • Standardiser les méthodes de déploiement • Uniformiser les pipeline de CI/CD • Définir le socle unifié d’hébergement des clusters kubernetes • Travailler avec les équipes plateforme pour définir les blocs communs unitaires de déploiement • Définir les bonnes pratiques sécurités Principales responsabilités & réalisations : • Définition des architecture kubernetes • Définition des architectures terraform • Définition des pratiques de sécurité au sein des comptes AWS • Définition des bonnes pratiques gitops Résultats / Impact : • Définition du Socle Minimal Applicatif Complet pour Kubernetes (SMACK): gestion des secrets, gestion du monitoring, gestion des coûts • Définition des architectures terraform pour le déploiement

- Transformed an existing data platform for a consulting firm into a modern, containerized solution, enhancing its commercial viability: - Migrated a VM-based Dagster data platform to a containerized architecture, improving portability and deployment consistency on OVH. - Developed Helm charts for essential platform components, including Dagster, PostgreSQL, and the monitoring stack. - Established a comprehensive CI/CD pipeline using GitLab CI, enabling: - Automated Docker image builds and security scanning. - Validation and testing of Helm charts. - Automated deployments to Kubernetes environments. - Implemented monitoring and alerting mechanisms using Prometheus and Grafana to ensure platform reliability. - Created detailed documentation and deployment guides to facilitate the commercial rollout of the platform

As data engineering at BNP (8th banking institution in the world) I was part of the 12 people team deploying dominodatalab at group level: - Developed and maintained Docker images tailored for a data science team, incorporating necessary tools and libraries, and implemented a robust CI/CD pipeline to automate the build, test, and deployment process, thereby enhancing team efficiency and ensuring consistent, reproducible research environments. - Managed Kubernetes clusters using Helm, ArgoCD, Kyverno, and Terraform to ensure the robustness and reliability of a data platform, and utilized Kibana for creating insightful dashboards to monitor system health and performance, thereby optimizing resource allocation and system uptime - Implemented Python scripts utilizing Kubernetes and Typer module to automate repetitive tasks on our data platform, significantly enhancing operational efficiency and reducing manual intervention.

OpenClassroom is the main french independant training platform allowing to everybody to learn and work in IT or even more. As mentor for this platform I accompanied a 30 years old student transitionning to IT: - Mentoring student every week to help him on his track - Working with the platform in order to evaluate the student track and adapt the training

As main cloud engineer in Datafab I was in charge of managing AWS Infrastructure - Designed and deployed a scalable web scraping system on AWS using Docker, ECS, and Terraform, with GitLab-CI for continuous integration and deployment, enabling efficient data extraction and enhancing the data-driven decision-making process. - Leveraged AWS API Gateway and Athena to ingest and query ERP data, providing a streamlined data pipeline that enabled data scientists to directly access and analyze the data using Athena, thereby facilitating more efficient data-driven decision making. - Implemented FinOps and governance strategies using CloudCustodian, successfully optimizing resource utilization and reducing overall cloud usage by 30%, resulting in significant cost savings and improved operational efficiency.

Integrated in Cloud Factory I worked on control tower deployment project in a team of 10 people and worked on day-to-day maintenance for website - Developed custom Terraform modules to automate the deployment of AWS Control Tower, streamlining the setup of multi-account AWS environments and enhancing governance, compliance, and operational efficiency. - Implemented infrastructure monitoring using Dynatrace, involving agent installation and alert configuration, to proactively identify and address system issues, thereby enhancing system reliability and performance. - Utilized Ansible to update and manage website changes, ensuring seamless deployments with zero downtime or impact to end users, thereby maintaining a consistent and high-quality user experience.

Formation opérationnelle a l'emploi : - Front-end: HTML/CSS/Bootstrap/Javascript - Back-end: Python/Flask - Base de données: SQL/NoSQL - Administration systemes: Docker/Ansible/Methodologie CI/CD - Autre: Terraform/Cloud/AWS

Skills: Anglais

Teaching how to use CATIA/ENOVIA (V5-V6) Software. Creation and modification of the training material. Support customers by creating methodology adapted to their specific needs.

Managing the 3D printers eco-system Advising the visitors of the lab for the manufacturing of their projects Helping StartUps implementing 3DExperience apps in their workflow

Worked as deputy director in holliday camp. This include but not limited to : - Recruiting the team by studying profile and making interview - Planing activity and making reservation for external providers - Interacting with the technical team (cook, maid...) - Leading everyday meeting in order to insure the respect of the pedagogic project and keep the team cohesion

Testing and crafting prototypes Realisation of manufacturing drawings/ Client drawing CAD modifications

Was in charge of leading the Nao project which aim to promote the IUT using the french robot NAO. During the 12 weeks I was in charge of : - Mechanical design and production of the part (3D printing, sheet metal, water cutting) - Programming arduino for the demonstration - Programming the NAO robot via the graphical interface/python Additional task : - Management and operating the 3D printers - Developing and hosting simplification activity about mechanics

Take part in the development of technical solution CAD/drafting on Solidworks