MLOps Engineer Job Description

About the role

Join us as an MLOps Engineer to build the backbone that gets machine learning models into production and keeps them healthy there. You will own model pipelines, deployment automation, feature stores, and monitoring, closing the gap between data science experiments and dependable live systems. We are looking for someone who loves automation and reproducibility, and who wants no manual step between a trained model and a stable production release.

Key responsibilities

• Automate model training, deployment, and retraining pipelines
• Build monitoring for model performance and drift
• Own the ML infrastructure and CI/CD for models

Responsibilities

• Build automated training and deployment pipelines with tools like Kubeflow, MLflow, or SageMaker
• Containerize models and serve them via scalable endpoints on Kubernetes or managed services
• Set up CI/CD for models, including automated testing, versioning, and rollback
• Design and maintain feature stores so training and serving use consistent features
• Implement monitoring for latency, throughput, data drift, and model performance decay
• Automate retraining triggers and reproducible experiment tracking
• Manage model registries and enforce lineage, approvals, and promotion between environments
• Optimize compute cost and autoscaling for GPU and CPU inference workloads
• Partner with data scientists to productionize experimental notebooks into robust services
• Build alerting and rollback runbooks so failing models are caught and reverted quickly

Requirements

• Bachelor's degree in computer science, engineering, or equivalent hands-on experience
• 3 or more years in software, DevOps, or data engineering, with direct MLOps responsibility
• Strong Python plus hands-on experience with Docker, Kubernetes, and CI/CD pipelines
• Proven experience deploying and monitoring machine learning models in production
• Familiarity with ML lifecycle tools such as MLflow, Kubeflow, or a major cloud ML platform