Whilst customer want to have dynamic environments for the data scientists to install packages/libraries whilst "experimenting" customer want to lock down their production environment. So that is going to be possible to have an environment where only the Admin can install packages at an environment level and therefore make it so that if a user runs a notebook which tries to install a new/updated library this will fail? This means customer can keep a solid grip on the production environment and just have a policy that whilst people can experiment with whatever environments they like, if it is going to production then it has to conform to "environment X".
Why is it useful?
|Who would benefit from this IDEA?||Data scientists would know exactly that the model they built is going to work in a production while deployment engineers would be able to govern the production environment at enterprise level.|
How should it work?
|Submitting Organization||F2F Sales|