[leksak]

Given https://github.com/jupyter/docker-stacks Docker might be appropriate. No matter which route you decide to take: make it easy. If you are using pipenv or poetry do so while offering a Makefile or documentation on what commands to run.

A Makefile with a decent "help" target would be perfect even if you do use Docker. I'd suggest looking at https://marmelab.com/blog/2016/02/29/auto-documented-makefil... to spare you some work with maintaining the help target.

[daww]

Thanks!