How Long Should It Take To Set Up a DS/ML CICD Pipeline on Azure Cloud?

Azure Devops is a fantastic set of Devops and CI/CD tools for Azure Cloud users. However, one concern I imagine many new Data/DS/ML engineers have on their mind is:

How long will it take me to set up a CICD pipeline for my data/ds/ml workflow on Azure Devops?

I was curious to explore this question.

What would it mean to answer this question?

Projects running on Azure, as on many other cloud providers, are broad, diverse, and can be composed of a myriad of technologies.

So I went with “Youtube Speed Runs”.

One basic assumption here is that nowadays, most of the initial contact developers have with new technologies happens by watching someone else press all the buttons.

Right away I hear you mention that “That’s not representative of a templated environment.. the team you are on probably already has a pipeline setup”. Yes that’s probably true, but imagine you had to debug a serious issue with the cicd pipeline, you probably will need to go through a set of tutorials on a different environment to learn how the pipes actually work.