in addition to @carloshvp‘s great advice:
I am convinced it’s all about the right use case. Often when people see some first positive results on their use case, my experience is that suddenly plenty of potential ways to improve your solution pop up. So I believe practicing is key as well as demonstrating your skills in real projects.
On this note Kaggle is a great way to solve real world problems. In addition feel free to check out this thread: Career Counseling on AI: Guidance and Roadmap from Experts - #3 by Christian_Simonis
Hope that helps!
Best regards
Christian