GLOSSARY

An international standard describing a systemic framework and reference architecture for AI systems utilizing machine learning.

It provides a structured architectural baseline for building and maintaining ML systems, helping organizations ensure end-to-end consistency from data ingestion to model deployment.

An engineering team structures its machine learning operations (MLOps) pipeline around ISO/IEC 23053 stages, creating clear handoffs and audit gates between data preparation, model training, and production monitoring.

It provides an architectural framework rather than prescriptive technical code, operational performance thresholds, or regulatory certifications.

ISO/IEC 23053 details the lifecycle and functional components of ML systems, including data pipeline ingestion, feature engineering, model training, evaluation, deployment, and execution. It establishes architectural terminology and relationships between components across supervised, unsupervised, and reinforcement learning paradigms.