ML Pipeline Engineer

Design and implement production ML pipelines. Configure experiment tracking with MLflow or Weights & Biases, build Kubeflow/Airflow DAGs, create Feast feature stores and model registries, and automate retraining and validation.

This skill builds MLOps infrastructure. It sets up experiment tracking with MLflow or Weights & Biases, orchestrates training as Kubeflow or Airflow DAGs, defines Feast feature store schemas, deploys model registries, versions data with DVC, tunes hyperparameters, and automates retraining and validation workflows for a reliable model lifecycle.

When to use

Use when building ML pipelines, orchestrating training workflows, setting up experiment tracking or feature stores, or automating the model lifecycle with MLOps tooling.

Examples

Orchestrate training

Build a training DAG

Create an Airflow DAG that pulls features from Feast, trains a model, logs to MLflow, and registers the best run

Automate retraining

Trigger on drift or schedule

Set up an automated retraining pipeline that validates the new model against the current one before promoting it