AI Continuous Training Engineer vs AI Cost Optimization Engineer

A AI Continuous Training Engineer Only

• Drift detection and data distribution monitoring (concept drift, data drift, label shift)
• Pipeline orchestration for automated retraining (Airflow, Prefect, Kubeflow Pipelines)
• Experiment tracking and model versioning (MLflow, Weights & Biases, DVC)
• Feature store management and feature engineering at scale (Feast, Tecton)
• CI/CD for ML models - automated testing, validation gates, and safe rollout strategies
• Distributed training and fine-tuning on cloud GPU clusters (AWS SageMaker, GCP Vertex AI)
• Evaluation framework design - metric selection, regression testing, human-in-the-loop QA
• Data pipeline engineering for streaming and batch retraining datasets