MLOps: GHWorkspace for Jupyter notebook

[pulkitaggarwl]

We can think of GitHub Workspace for Jupyter Notebook an environment for Notebooks where in a single click an environment will be spun up for the Data Scientist allowing them to run the notebook without them having to install any dependencies.

Targeted Customer Journey:
Customer opens a ML application in "GitHub workspace for Jupyter" which creates a fork
He can run the notebook without having to install any dependencies.
Makes changes to the notebook and data pipeline as needed and iterate
Executed the model and plot metrics.
Saves the updated model to forked branch.

[marvinbuss]

This is a really important and fundamental feature that is missing.
@awmatheson we have been talking about this a couple of months ago.

[marvinbuss]

@pulkitaggarwl

• Should we wait for codespaces?
• Large overlap with codespaces. Will it be supported in codespaces?
• Synch with AML team: Can the experience be managed by GitHub and the infrastructure be maintained by AML? Can we build Jupyter extensions for AML?