Model Registry Service

The Model Registry Service is where your ML models are safely stored, tracked, and shared as Docker images. It handles all the complexity of working with container registries, letting developers and MLOps teams focus on building, testing, and deploying models.

It ensures models are versioned, consistent, and ready to scale, with automatic synchronization between local and remote copies. Security and quality are built in through automated validation and vulnerability scans. Role-based access controls and metadata management make governance and collaboration simple.

Integrated with harbor, the service combines secure container storage with automated workflows, giving teams a reliable, enterprise-ready platform for managing models from development to production.

The following topics are available:

• Model Registry Service Requirements
• Model Registry Service Design
• Model Registry Service Runbook (Happy Path)