Overview
We are seeking a Mid-Level MLOps Engineer to build, operate, and evolve our Kubeflow-based ML platform on Azure. This role focuses on enabling reliable, scalable, and cost-efficient ML workflows by designing CI/CD pipelines, managing Kubernetes-based ML infrastructure, improving platform observability, and supporting MLE and Data Science teams across the model lifecycle. The idóneo candidate is hands‑on, comfortable working across infrastructure and ML workflows, and motivated to operationalize best practices in MLOps.

Responsibilities
Platform & Infrastructure:

Deploy, configure, and operate Kubeflow components on Azure Kubernetes Service (AKS)

Support Kubernetes workloads for training, inference, and batch pipelines

Manage container images, registries, and ML runtime environments

Assist with Kubeflow and Kubernetes upgrades under senior guidance

CI/CD & Automation:

Build and maintain CI/CD pipelines for ML workflows and platform services

Automate model training, validation, and deployment pipelines

Implement reproducibility and versioning for data, models,



and pipelines

Observability & Reliability:

Implement logging, monitoring, and alerting at the platform level

Diagnose and resolve workflow, pipeline, and infrastructure failures

Support SLAs and reliability objectives for ML platforms

Collaboration & Enablement:

Work closely with MLEs and Data Scientists to onboard workflows onto Kubeflow

Provide best practices, templates, and documentation for ML teams

Collaborate with Infra and Security teams on access control and compliance needs

Cost Awareness & Optimization:

Assist with collecting and reporting costs at Kubeflow namespace or workflow level

Identify optimization opportunities related to compute usage and scheduling

Qualifications

3-6 years of experience in MLOps, DevOps, or Platform Engineering

Hands-on experience with Kubeflow, Kubernetes, Terraform (IaC), and containerized ML workloads

Strong experience with Azure cloud serv

📌 MLOps Engineer (Madrid)
🏢 PepsiCo
📍 Madrid