Model Registry
A model registry is a versioned catalog that tracks every trained ML model artifact, along with its metadata, training data lineage, evaluation metrics, approval status, and deployment stage (staging, production, archived).
Full definition
The registry is the system of record for "which model is in production right now, who approved it, and what data trained it." Common implementations include MLflow Model Registry, SageMaker Model Registry, and Vertex AI Model Registry. Under the EU AI Act, high-risk AI providers must keep technical documentation and post-market records — a registry is the natural place to anchor that obligation.
Why it matters
Without a registry, model rollbacks become forensic exercises and audits become spreadsheet archaeology. A registry makes "promote model v37 to production" and "roll back to v36" idempotent CLI commands, and it gives risk and compliance teams a single page to ask "what is running, on what data, approved by whom."
Example
A bank promotes credit-scoring v42 to production through the registry; an alert fires two days later, the on-call engineer rolls back to v41 in 90 seconds and the registry preserves the full incident trail.
Related terms
- MLOpsMLOps is the discipline of operating machine-learning and AI systems reliably in production — covering data pipelines, model training, deployment, monitoring, drift detection, governance, and incident response — analogous to DevOps for traditional software.
- AI AuditAn AI audit is a structured, evidence-based examination of an AI system or AI program against defined criteria — covering training data, model, deployment context, monitoring, and governance — performed by an internal team, an external firm, or a regulator.
- AI GovernanceAI governance is the framework of policies, roles, controls, and processes an organization uses to ensure its AI systems are lawful, safe, fair, accountable, and aligned with business intent — across the full lifecycle from problem framing to retirement.
- Model CardA model card is a short, structured document that describes an AI model's intended use, performance, limitations, training data, evaluation results across demographic and operational subgroups, and known failure modes — the AI equivalent of a nutrition label.
Source & further reading
Primary source: MLflow Documentation — Model Registry (2024).
Citation policy: this entry is part of the AIDOLS AI Implementation Glossary and may be quoted for research, journalism, and education with attribution to aidolsgroup.com/en/glossary/model-registry/.