Fantastic overview!

Agree that the MLOps is just getting started. Imho, the whole data science is.

The data industry need to deliver on the expectations of the businesses. Building an ML model is one thing, having it make decisions automatically, with proper monitoring and QA is a different game. Let’s hope we succeed.

I also think that not every company need to build their own ML models, but most of them will be using ML models. So having the right technology and best practices for that will be key.