Marco Braun
Machine Learning Engineer @OVHcloud
Signup · Get unlimited contacts
WORK HISTORY
Machine Learning Engineer @OVHcloud
Munich, DE
Building and operating the platform that runs customer AI workloads on OVHcloud — AI Training jobs, Notebooks, AI Deploy, and AI Endpoints — across four regions in Europe and North America.What I work on:• Control plane engineering in Go — REST APIs, Kubernetes controllers, and microservices that schedule customer ML workloads onto GPU clusters• ovhai CLI in Rust — the primary interface ML engineers use to manage jobs, notebooks, models, data buckets, and registries• Production on-call across 4 regions: incident response for Ceph storage, GPU node failures, etcd degradation, Argo workflow issues, and customer-facing platform alerts• Cross-team collaboration with frontend (Vue.js), infrastructure (Terraform/Kubernetes), and SRE on platform reliability and developer experienceStack: Go, Rust, Kubernetes, Ceph, Argo Workflows, Prometheus, Grafana, gRPC, REST.
EDUCATION
Hochschule Fresenius
Bachelor, Business Psychology
ABOUT MARCO BRAUN
I work on the cloud infrastructure that powers OVHcloud\'s AI suite — Training, Notebooks, AI Deploy, and AI Endpoints — running customer ML workloads on GPU clusters across four regions in Europe and North America (Gravelines, Beauharnois, Kepler, US-East-VA). My day-to-day spans the control plane (Go microservices, Kubernetes operators, REST APIs) and the Rust CLI (ovhai) that ML engineers use to manage their jobs, notebooks, and data.I\'m drawn to the systems engineering behind running GPUs at scale: multi-tenant scheduling on Kubernetes, distributed storage on Ceph, Argo Workflows for cache and lifecycle management, and the on-call discipline that keeps customer training jobs running through node failures, etcd degradation, and storage incidents. On-call rotation across all four regions means I see what actually breaks in production AI infrastructure — and what it takes to keep it boring.Background: 5+ years of backend engineering. Before OVHcloud I was at Exxeta migrating enterprise document and e-commerce systems to serverless AWS architectures with Node.js, DynamoDB, and Kafka. Comfortable across Go, Rust, Python, and TypeScript, with a particular interest in the systems work that turns raw GPUs into a reliable platform. Outside of work: gym, basketball court, and a deep love for Italy and its pizzas. Let\'s connect if you want to chat about infrastructure, GPUs, sports, or the perfect Neapolitan crust.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.