Machine Learning Systems and Infrastructure Developer

Research & Development

Change the future with us.

We are looking for dedicated and talented people who solve ever-changing challenges, customer needs, and questions from colleagues with clever concepts and creativity. We embrace change and work with curious minds re-inventing the future of work. Join us and let us focus together on what’s truly important: making lives better with current ideas and the latest technology around the world.

About the role:

Are you excited by the engineering work that turns model development into repeatable, governed production rollout?

Do you enjoy building the infrastructure behind training jobs, GPU environments, experiment tracking, model registries, evaluation gates, and controlled deployment of robotics AI models?

We are seeking a Machine Learning Systems and Infrastructure Developer to build the training infrastructure and model rollout layer for Physical AI. This developer will provision and operate ML environments, automate training and evaluation pipelines, manage model artifacts and registries, and define the promotion path from trained model to validated runtime candidate. The role connects robotics data, partner VLA/RFM work, model evaluation, release gates, and runtime deployment without owning robotics data semantics or the final runtime integration layer alone.

What will you do?

- Provision and maintain ML training environments for robotics AI workloads across local, edge, on-premise, and cloud-backed infrastructure where applicable.
- Build repeatable training, fine-tuning, evaluation, and model-packaging pipelines for VLA, VLM, perception, and robot-learning workflows.
- Manage model artifacts, model registries, dataset references,



experiment metadata, lineage, approvals, promotion states, and rollback paths.
- Automate GPU resource allocation, job scheduling, environment reproducibility, dependency management, and capacity tracking.
- Define model rollout workflows from training output to validated runtime candidate, including evaluation gates, release notes, traceability, and operational handoff.
- Partner with the Senior AI/ML and Robotics Platform Engineer on dataset readiness, evaluation semantics, and model-quality criteria.
- Partner with ML inference, Edge Runtime, AX integration, QA, and DevOps engineers on packaging, deployment readiness, observability, and release automation.

What you will bring:

- Bachelor's or Master's degree in Computer Science, Machine Learning, Robotics, Computer Engineering, Software Engineering, Electrical Engineering, or a related field.
- 3+ years of Strong experience building ML platforms, training infrastructure, MLOps systems, model lifecycle tooling, or production AI infrastructure.
- Strong programming skills in Python and practical experience with containers, Linux, automation, APIs, configuration management, and reproducible environments.
- Experience with training orchestration, job scheduling, GPU provisioning, artifact management, experiment tracking, model registries, or release pipelines.




- Practical experience with ML frameworks and infrastructure such as PyTorch, Hugging Face tooling, Ray, Kubernetes, Slurm, MLflow, Weights & Biases, Docker, CUDA, or comparable systems.
- Understanding of GPU capacity, storage, networking, data locality, dependency isolation, cost controls, and failure recovery for training workloads.
- Experience connecting datasets, metadata, evaluation outputs, model artifacts, and deployment packages through traceable workflows.
- Ability to define model promotion and rollback processes with clear gates for quality, reproducibility, security, and runtime readiness.
- Experience partnering with data engineering, ML research, robotics, DevOps, QA, and runtime teams to make model development operational.
- Strong troubleshooting skills across training jobs, data access, dependencies, GPU infrastructure, model artifacts, and rollout automation.

What sets you apart:

- Experience with VLA, VLM, robotic foundation models, robot learning, imitation learning, reinforcement learning, or multimodal training workflows.
- Experience operating GPU clusters, shared training platforms, high-volume data pipelines, or hybrid on-premise and cloud ML infrastructure.
- Experience with distributed training, multi-node GPU workloads, or training orchestration frameworks such as Ray, Slurm, Kubernetes, Kubeflow, or comparable systems.
- Experience designing portable ML infrastructure across on-premise, cloud, or multicloud environments, including data locality, reproducibility, access control, and cost management.
- Familiarity with dataset versioning, data lineage, model cards, evaluation reports, approval workflows, and audit-ready mo

📌 Machine Learning Systems and Infrastructure Developer (Kitchener)
🏢 Siemens
📍 Kitchener