Martin Roch-Neirey
Networks, systems, cloud & AI infrastructure
M.Eng. student seeking an internship · between January 6 and May 2, 2027 · full-time, 450 hours minimum
Master's student whose studies have focused mainly on networks and systems. I have also worked on GPU infrastructure and HPC (Slurm and Kubernetes clusters), as well as green computing, through a Bachelor's thesis on carbon-aware scheduling. Curious and versatile, I am eager to learn alongside experienced teams.
Experience {
Co-founder, Cloud & Kubernetes Infrastructure
- Designed, built and operate the company's production infrastructure (managed Kubernetes on Scaleway, Cloudflare), fully defined as code with Terraform, so it can be rebuilt and audited.
- Set up GitOps continuous deployment (ArgoCD) for the development team, with development, staging and production environments; documented decisions and runbooks for the team.
- Secured the platform with Zero Trust: internal tools reachable only through Cloudflare Tunnel and SSO, origin locked to Cloudflare (Authenticated Origin Pulls, mTLS).
HPC & GPU Infrastructure, CS Department
- Rebuilt the Kubernetes cluster (3 control-plane, 15 workers, bare-metal and vSphere VMs, cloud bursting): NVIDIA GPU Operator, Ceph CSI, Rancher, ingress controllers per access policy.
- Built and operated the school's first Slurm cluster: NVIDIA GPUs (GH200, A40, A6000), CUDA, vGPU on vSphere, three storage spaces, automated deployment and monitoring.
Scientific Collaborator
- Audited the institute's heterogeneous GPU cluster usage and recommended efficiency gains.
Lecturer, Telecommunications & IP Networks
- Taught the Telecommunications and IP Networks course to future engineers.
Java Developer & Project Management
- Project management (meetings, specifications, decisions) for a network of up to 2,500 concurrent and 1M unique players; recruited and led a 20-person moderation team.
- Java minigames and contributions to internal systems (API, part of the anti-cheat); GitLab, Jenkins, ELK.
System Administration Intern
- 12-week internship: designed and built a secure CI/CD pipeline deploying an application to Kubernetes, with code, container image and runtime security analysis.
/* Projects & research */
Benchmarking cluster compute capacity
Terraform, Ansible and HPL to compare vSphere, OpenStack and Kubernetes clusters.
Bachelor's thesis, EU project FLUIDOS
Carbon-aware multi-cluster Kubernetes placement; co-authored paper under review (IEEE journal).