ML Engineer - Infrastructure at Flexion Robotics
Switzerland
Posted
on-site · full_time · Visa sponsorship
Apply for this role at Flexion Robotics
Responsibilities:
- Architect, provision and operate large-scale GPU training clusters (cloud and/or on-prem).
- Select and integrate schedulers, orchestration and monitoring tooling (Slurm, Kubernetes/KubeRay or similar).
- Improve training throughput and utilization: profiling, DDP/FSDP/NCCL tuning and pipeline optimisation.
- Support AI teams to optimise code, pipelines and data workflows for multi-node training.
- Drive compute strategy: capacity planning, cost optimisation and multi-cloud trade-offs.
- Raise engineering standards for testing, reliability, IaC and operational runbooks.
Requirements:
- Degree in CS, EE or equivalent practical experience and significant industry experience.
- Hands-on experience running large-model training (multi-node GPU clusters) and bringing up GPU infrastructure.
- Strong Python skills and working knowledge of PyTorch.
- Deep understanding of distributed training concepts (DDP, FSDP, NCCL) and performance profiling.
- Experience with cloud platforms or large-scale on-prem GPU fleets and job schedulers (Slurm, Kubernetes, KubeRay or similar).
- Familiarity with infrastructure-as-code, monitoring and reliability practices; ability to collaborate with AI and infra teams.