Story (Required)

As a Platform Engineer, I should be able to provide documentation about each model in my organization that includes: Basic model information, its purpose, its license, and its type.

As an Application Developer, I should be able to easily view that information from within a model’s Resource catalog entry. The information should be clear to understand and written in a consistent manner.

Background (Required)

As part of DEVAI-147, a way to model AI assets (models, model servers) in a Backstage catalog will be implemented. This issue will focus on providing a consistent way to document basic model metadata, that isn’t covered by fields in Backstage Resources. This information may include basic model overview, its purpose, its license, its type, and where to download it.

One or more techdocs (in the form of a markdown file) should be provided that provides that necessary information for the model. It should be formatted in a way such that its template can be reused, regardless of the model that it’s for.

Out of scope

• Any documentation needed for the model server

Approach (Required)

• Define a techdoc format providing information about models in an organization
• Document structure and how users can create techdocs from it.

Dependencies

Acceptance Criteria (Required)

• Specification provided that implements the Model Catalog API (e.g. Google Doc or Git repo)
• Every, eligible non-techdoc related field in the Model Catalog API is implemented

Done Checklist