Goal- Build upon integration between Model Registry (MR) and Quay

• Enable metadata indexing to support queries such as retrieving all Model Card information.
• Ensure tag immutability for ML models to support downstream integrations 
• Support seamless authentication between OpenShift AI and Quay using a keyless approach.

Objectives

• Seamless UI for Model Navigation: Improve Quay’s UI to provide an intuitive browsing experience for ML models.

• Searchable Model Metadata: Indexed metadata enables efficient search and retrieval of ML models and Model Cards.

• Stronger Security & Governance: Tag immutability ensures ML model integrity and compliance.

Acceptance Criteria

Seamless UI for Model Navigation

• Users can visually browse and interact with ML models in Quay without relying on manual queries.

• UI enhancements align with Model Registry for a unified experience.

Searchable Model Metadata

• Users can issue structured queries to retrieve Model Card metadata from Quay.

• Model metadata is fully indexed, searchable in the UI, and filterable based on attributes such as model type, version, and last updated date.

Tag Immutability for ML Models

• Tags for ML models are immutable to ensure governance and integrity.

• Users cannot delete or modify ML model tags once published.