skills.rest
martinholovsky
martinholovskyCOMMUNITY
Explore all skills by this author →
𝕏@martinholovskyGitHub

sota-ml-engineering

Provide rules and guidelines for building and auditing production-ready machine learning systems.

12|2|Updated Jun 17, 2026
Search Tags:#MLOps#reproducibility#data leakage#data versioning#model governance#ML system architecture#ML security
One-click install
npx skills add https://github.com/martinholovsky/SOTA-skills --skill sota-ml-engineering
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sota-ml-engineering
Source: https://github.com/martinholovsky/SOTA-skills/tree/main/skills/sota-ml-engineering
Command: npx skills add https://github.com/martinholovsky/SOTA-skills --skill sota-ml-engineering

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides a comprehensive set of rules and guidelines for building and auditing machine learning systems, ensuring they are robust, secure, and ready for production.

Core Features & Use Cases

  • ML System Architecture: Defines best practices for designing ML systems with a focus on the model as part of a larger system.
  • Data and Features: Offers guidance on data and feature management, including data leakage, train/serve skew, and versioning.
  • Training and Experimentation: Provides reproducibility and tracking rules for training and experimentation processes.
  • Evaluation and Validation: Ensures models are evaluated against business objectives and are production-ready.
  • Deployment and Serving: Covers model packaging, serving patterns, and operational considerations.
  • Monitoring and Drift: Focuses on ongoing monitoring of model performance and data drift.
  • Security and Governance: Offers guidelines for security best practices and compliance with regulatory standards.
  • Use Case: Utilize this Skill to build a machine learning system that is reproducible, secure, and compliant with industry standards.

Quick Start

Run the 'build-ml-system' command to initiate the process of building an ML system according to the sota-ml-engineering rules.

Frequently Asked Questions about sota-ml-engineering

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent data leakage when building machine learning systems?▼

ML system architecture design treats the model as a component within a larger system, requiring strict boundaries for data management, deployment, and monitoring. This approach ensures robust production-readiness and operational stability.

What is the best way to audit ML systems for security and governance?▼

To audit ML systems for security and governance, verify model packaging, serving patterns, and operational guidelines against established compliance standards. This ensures robust security and regulatory adherence across the deployment pipeline.

How do I set up reproducibility for ML training and experimentation?▼

Reproducibility in ML training and experimentation requires enforcing strict tracking rules for data versioning and feature management. Documenting these processes ensures that all training iterations can be reliably replicated and audited for production readiness.

How do I monitor machine learning models for data drift in production?▼

Monitoring machine learning models for data drift requires implementing ongoing performance tracking rules within your ML system architecture. This detects deviations in data features and model behavior, ensuring sustained production readiness.

Does this approach to MLOps work for evaluating models against business objectives?▼

Yes, this MLOps approach works for evaluating models against business objectives by providing specific validation rules. It ensures models are technically robust while remaining aligned with production-ready business goals.

Related Skills

Explore More Software Engineering Skills→
YeomanYeYeomanYe

project-rules-design

Restructure project engineering rules into domain-based index.md and rules.md files.

Community
Advanced
KentoShimizuKentoShimizu
7

mlops-monitoring-drift

Detect data drift, concept drift, and quality degradation in production ML pipelines.

Community
Advanced
informatico-madridinformatico-madrid
1

mlflow

Manages ML lifecycle with experiment tracking, model registry, and deployment tools for TensorFlow, PyTorch, Scikit-Learn.

Community
Advanced
Sir-chawakornSir-chawakorn
9

monitor-ml-drift

Monitor production ML models for data drift and performance decay using statistical tests.

Community
Advanced