Can I use this for cloud-specific ML services?

Yes, it includes integration patterns for managed platforms like AWS SageMaker, Google Vertex AI, and Azure ML.

What orchestration tools does this skill support?

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

How does it handle model validation before deployment?

It provides frameworks for A/B testing, performance regression detection, and comprehensive pre-deployment checklists to ensure model quality.

Does it help with data versioning?

Absolutely, the skill covers data lineage and versioning strategies using tools like DVC and TFX to ensure complete pipeline reproducibility.

ML Pipeline Workflow

Name: ML Pipeline Workflow
Author: HermeticOrmus

byHermeticOrmus

データサイエンスとML

Builds and orchestrates end-to-end MLOps pipelines from data preparation through model training to production deployment.

概要

This skill provides comprehensive guidance for designing and implementing production-grade machine learning pipelines using MLOps best practices. It covers the entire lifecycle, including data ingestion, feature engineering, automated training orchestration, rigorous model validation, and reliable deployment strategies. Whether you are using Airflow, Kubeflow, or cloud-native tools like SageMaker, this skill helps automate reproducible workflows, ensure data lineage, and maintain model performance through observability and automated rollback mechanisms.

主な機能

End-to-end DAG orchestration patterns for complex ML workflows
0 GitHub stars
Production deployment strategies including canary and blue-green releases
Comprehensive model validation and performance regression testing
Automated model training and experiment tracking integration
Data validation and feature engineering pipeline design

ユースケース

Building reproducible automated training pipelines for new ML models
Designing scalable model serving architectures with automated rollbacks
Implementing data drift detection and continuous retraining workflows

概要

主な機能

End-to-end DAG orchestration patterns for complex ML workflows
0 GitHub stars
Production deployment strategies including canary and blue-green releases
Comprehensive model validation and performance regression testing
Automated model training and experiment tracking integration
Data validation and feature engineering pipeline design

ユースケース

Building reproducible automated training pipelines for new ML models
Designing scalable model serving architectures with automated rollbacks
Implementing data drift detection and continuous retraining workflows