Bridging Embedded Systems Expertise to the World of Machine Learning

I’m a computer engineering graduate and come from a traditional embedded systems background, with knowledge of microcontrollers, computer architecture and operating systems. Is knowledge of C and C++ programming, Linux networking, memory management , multithreading, synchronization, interrupts etc useful in ML engineering. Are subjects like distributed systems, compiler optimizations (using LLVM), parallel computing etc going to be useful or are they heavily going to be automated as well by AI? In other words, is computer engineering always going to required to scale ML systems and be evergreen? Are people in ML engineering using these skills in their work everyday? Thank you.