Tips to break into the Machine Learning industry

Hello blinders,

I have been working as an Embedded Software engineer for the past one and a half years. I got into this field right out of college. But nowadays, I am getting bored while working my day to day job. I also do not see much development/innovation in this field in the near future which might mean I would not be experiencing fast paced growth in terms of job satisfaction, TC, challenges etc.
I always had interests in the ML field as I read a lot of books and watched a lot of documentaries related to AI but I never actually took the initiative to actually implement real world applications through personal projects as I did not have the theoretical knowledge of ML models like regression, clustering etc.
So my question to the Blind community is:
1) How do I transition from Embedded SWE to ML expert? For e.g Do I apply to Embedded positions that are somewhat related to ML like the Tesla Autopilot team that works with the FSD chip? Or is it advisable to directly apply to entry level ML positions.
2) I heard that you require a very specialized ML background like having a MS in Machine Learning or a PhD to break into this field. Is that true? I have a MS in Computer Science so I do not have that specialized formal knowledge.
3) What ML field/work should I look into now to start learning and applying my knowledge on different projects for potential employers? I am good with C/C++, Python as I use these 3 for day to day work.
4) Any resources that I should start using now in order to gain the basic knowledge of programming ML models?
5) Once I have the knowledge what type of positon should I apply to based on my experience as an engineer? For eg entry to mid entry level etc.

TC: 135K
YOE: 1.5