Machine Learning Training Infrastructure Engineer

Machine Learning Training Infrastructure Engineer

Logitech

Lausanne, Switzerland

The Role:

We are seeking an ML Training Infrastructure Engineer to elevate the engineering practices across our Machine Learning and EdgeAI teams and improve their velocity. In this role, you will act as a shared champion of DevOps / MLOps excellence. Your mission is to drive excellent model training by designing the reusable scaffolding (templates, CI/CD pipelines, and data management workflows) that allows our ML engineers to work faster, smarter, and with absolute consistency. A key focus of this role is optimizing training pipelines to fully leverage our High Performance Computing (HPC) infrastructure, ensuring scalability and performance. You are the interface with our Digital Office (who manage core AWS/HPC resources and MLOps infrastructure) and the applied ML teams, ensuring knowledge transfer and engineering best practices are seamlessly integrated.

Your Contribution:

Be Yourself. Be Open. Stay Hungry and Humble. Collaborate. Challenge. Decide and just Do. Share our passion for Equality and the Environment. These are the behaviors and values you’ll need for success at Logitech. In this role you will look after:

• Infrastructure rollout & onboarding: Lead infrastructure migration and adoption by guiding ML engineers to new shared compute environments, optimizing developer workflows to reduce friction, and providing comprehensive documentation and hands-on training to ensure team upskilling;
• Enablement & Knowledge Transfer: Act as the internal consultant and coach for ML teams for High Performance Computing (HPC) optimization. Run workshops, pair-program on pipeline bottlenecks, and document modern MLOps standards to upskill the engineering org and facilitate knowledge sharing;
• Standardization & Templating: Design and maintain "cookie-cutter" project templates. Ensure any ML engineer can spin up a new repository that is automatically compliant, structured, and integrated with data/artifact versioning;
• CI/CD & Continuous Training (CT): Build and maintain robust automation pipelines. Design the triggers and workflows that automate data validation, model retraining, and evaluation when new data arrives or performance baselines change;
• Compute & Code Optimization: Partner with ML engineers to profile and optimize training pipelines and local/edge inference code in pair work, ensuring high efficiency on local host computers and specialized compute resources;
• Digital Office Liaison: Act as the primary technical point of contact to translate the Digital Office’s AWS policies, high-efficiency compute setups, and infrastructure realities into practical, developer-friendly workflows for the ML teams. You are the product owner of the MLOps toolchain making sure the requirements of the ML team are implemented as tools by the Digital Office;
• AI Governance & Data Stewardship: Establish and enforce frameworks for ethical AI, model transparency (via model cards), and data privacy compliance. Act as the steward of our data ecosystem by aligning processes for data inventory, lifecycle management, and accessibility, ensuring that high-quality, compliant data is readily available for training and inference across all projects.

Key Qualifications:

For consideration, you must bring the following minimum skills and experiences to our team:

• Education: PhD or MSc in Computer Science, Computer Engineering, or a closely related field;
• Experience: 5+ years of experience in MLOps, DevOps, or Software Engineering with a heavy focus on developer tooling, automation, platform enablement and pipeline optimization. Hands-on experience designing, building, or maintaining large-scale ML training infrastructure. You have trained ML models on HPC platforms as part of your prior experience and have been acquainted with the different elements of the stack, from data management to driver management;
• The "Leadership" Mindset: Proven track record of mentoring teams, establishing best practices, and creating clean documentation that engineers actually enjoy using;
• Data & Workflow Tooling: Strong, practical experience with data versioning, experiment tracking tools, and modern CI/CD platforms (e.g., GitLab CI, GitHub Actions);
• Software & Packaging Mastery: Deep proficiency in Python and PyTorch, with experience running training workloads on GPUs/TPUs. Strong understanding of containerization (Docker) and packaging models for distribution to edge devices or host computers. Strong software engineering practices: code quality, design reviews, testing, observability, CI/CD;
• Familiarity with Infrastructure: Solid understanding of cloud concepts (AWS) to effectively collaborate with the Digital Office, even though cloud deployment is out of scope;
• Systems & Performance Optimization: Strong understanding of profiling, operating systems (OS), caching mechanisms, and hardware (HW) architecture. Solid understanding of distributed systems concepts (parallelism strategies, fault tolerance, synchronization).