It's generally agreed that 80% of the effort in data projects goes into things that "data scientists" don't want to do.
It seems like everybody and his sister has products and services that address the 20% of the work that people want to do, but even if you can automate and accelerate that work way it can at best have a small impact on the cost and capabilities of data projects.
That of course means following a "path less followed" and I can just say it is a lot of work just to educate people that something can be done about it. I am trying here