Description of problem:

There are a lot of questions around how a model will be served. Who will have access to it? What kind of traffic is it expected? How quickly should it respond to requests? Do we care about high availablity? How do you define the resources for the deployment? Is the model available internally and externally?

Recommendation:

If a data scientist is also deploying a model, they will need to have strong defaults on what to select. Will need to document this process for customers who do not know this process already or if the data scientist is also deploying the model.

Additional info:

Plan to improve when have model serving. Have public URL to walk users through process. https://redhat-scholars.github.io/rhods-od-workshop/rhods-od-workshop/index.html