How does it handle model deployment?

The skill covers various serving patterns and automated deployment strategies such as canary releases, blue-green deployments, and shadow launches.

What orchestration tools does this skill support?

It provides patterns and templates for popular orchestration tools including Apache Airflow, Dagster, Kubeflow Pipelines, and Prefect.

Can I use this for data validation?

Absolutely. It incorporates data quality checks and validation frameworks like Great Expectations to ensure data integrity throughout the pipeline.

Does this skill help with experiment tracking?

Yes, it includes integration guidance for industry-standard tracking tools like MLflow, Weights & Biases, and TensorBoard.

ML Pipeline Workflow

Name: ML Pipeline Workflow
Author: HermeticOrmus

byHermeticOrmus

•

Data Science & ML

Orchestrates end-to-end MLOps pipelines from data preparation through production deployment and monitoring.

About

This skill provides a comprehensive framework for building and managing production-grade machine learning pipelines within Claude Code. It guides users through the entire MLOps lifecycle, including automated data validation, training orchestration, experiment tracking, and robust deployment strategies like blue-green or canary releases. By offering standardized patterns for DAG-based orchestration and integration with industry-standard tools like Airflow and MLflow, it ensures reproducible workflows and reliable model delivery for data scientists and ML engineers.

Key Features

2 GitHub stars
End-to-end DAG orchestration for Airflow, Kubeflow, and Dagster
Automated data validation and feature engineering pipelines
Integrated experiment tracking and hyperparameter management
Model validation frameworks with performance regression detection
Advanced deployment patterns including Canary and Blue-Green strategies

Use Cases

Building a reproducible training workflow for deep learning models
Automating model retraining and redeployment triggered by data drift
Designing scalable data preparation and feature engineering pipelines

About

Key Features

2 GitHub stars
End-to-end DAG orchestration for Airflow, Kubeflow, and Dagster
Automated data validation and feature engineering pipelines
Integrated experiment tracking and hyperparameter management
Model validation frameworks with performance regression detection
Advanced deployment patterns including Canary and Blue-Green strategies

Use Cases

Building a reproducible training workflow for deep learning models
Automating model retraining and redeployment triggered by data drift
Designing scalable data preparation and feature engineering pipelines