AI Platform Engineer — MLOps & Cloud Infrastructure

As AI Infrastructure Engineer within the CDO AI Services team, you design, build, and maintain the infrastructure that powers AI-driven solutions across the organization. You own the scalability, reliability, security, and cost-efficiency of a modern AI platform on Azure, GCP, or Databricks, including GPU-accelerated compute. Working at the intersection of cloud engineering, DevOps, and ML engineering, you enable data scientists and developers to train, validate, deploy, and monitor models reliably. A senior, hands-on role in a complex, high-scale environment where ownership and proactivity are essential.

• Design, build, and maintain scalable, robust AI platform infrastructure.
• Architect and manage cloud-native infrastructure on Azure, GCP, or Databricks.
• Leverage cloud-native services: compute, storage, networking, serverless, managed AI.
• Build and maintain ML pipelines for ingestion, training, evaluation, and deployment.
• Implement Infrastructure as Code with Terraform or Ansible.
• Design and maintain CI/CD pipelines for infrastructure and ML workloads.
• Integrate MLflow, Kubeflow, or NVIDIA Triton for experiment tracking and model management.
• Enable MLOps: model versioning, reproducibility, and lifecycle management.
• Monitor infra and ML workloads (logging, monitoring, alerting); define SLOs/SLIs.
• Optimize cloud cost efficiency (FinOps) and implement secure cloud architectures.
• Collaborate with data scientists to productionize models; drive automation-first practices.

• 7–10 years in infrastructure engineering, ideally AI/data platforms.
• Hands-on with Azure, GCP, or Databricks.
• GPU-accelerated compute environments.
• AI tooling: NVIDIA Triton, Kubeflow, or MLflow.
• Expert IaC: Terraform or Ansible.
• Expert containerization/orchestration: Docker and Kubernetes.
• Proven CI/CD pipeline design; ML pipelines (train/validate/deploy/monitor).
• Monitoring, logging, alerting for infra + ML workloads.
• Bachelor's or Master's in CS, Engineering, or related.