Thomas Moreau

- Modernized CI/CD infrastructure by migrating from Buildbot to Woodpecker CI with custom Helm charts, reducing deployment complexity and improving pipeline scalability - Led technical evaluation and migration from Gerrit to Gitea, streamlining code repository management and developer workflows - Deployed and managed production Kubernetes clusters (kubeadm, k3s) in high-availability configurations using Ansible automation - Architected scalable AI/ML infrastructure with GPU orchestration via Nvidia Operator, managing workflows across multiple GPU types (2080Ti, Quadro RTX 6000, RTX 4090Ti) with CUDA optimization and ClearML integration - Implemented auto-scaling solutions for AI/ML training agents, optimizing resource allocation and computational efficiency - Developed custom Golang scripts to enhance Kubernetes autoscaling and pod lifecycle management - Managed application versioning and deployments using Helm, helmfile, and integrated CI/CD pipelines - Optimized CI pipeline execution times by implementing shared SMB installation caches for Python dependencies - Implemented automated Kubernetes backup solutions using Velero, restic, and Longhorn with self-hosted MinIO S3 storage - Built blue-green deployment demo platform for AI solutions in close collaboration with development teams - Developed production-grade Python applications using modern tooling (Poetry, docker-compose, devContainers) for complete project lifecycle management - Designed and architected scalable, distributed systems in Python for production environments - Conducted advanced Python code reviews and championed best practices to ensure high code quality - Applied Agile methodologies (Scrum, Kanban) to drive iterative development and team collaboration

- Collaborative project with Hear and Know: Designed a transport bag for transplant organs using IoT with embedded cards and sensors. - Developed a machine learning algorithm to detect dangerous transport situations. - As DevOps Engineer: Managed deployment on IBM Cloud, integrated MQTT/LTS protocols, and ensured system reliability. - Technologies: IBM Cloud, Python/C++, Embedded Electronics/IoT, Machine Learning, MQTT/LTS, Arduino IDE.

- Internship in Digital Services (DevOps): Set up and experimented with server virtualization for applications. - Created an optical recognition service for identity documents. - As DevOps Engineer: Managed containerization with Docker, message queuing with RabbitMQ, and integrated OCR using Tesseract. - Technologies: Docker, RabbitMQ, Tesseract, SSH, Python/C++