I ask because I am myself an ML engineer, and I got interested if I am capable of creating a company out of scope of my main expertise.
I believe you can be successful without necessarily knowing a lot about the industry in the beginning. However, you need to talk to people in that industry about their problems.
You will only succeed if you solve a real problem at some point in time.
Say, you want to solve a problem in the restaurant space. Maybe even an idea that can be solved by you data skills. How do you know what restaurant owners will actually need? What kind of problems do they really have. Do they have similar problems or do they have unique problems.
The more your idea/company becomes deep tech, the more you need to rely on people to make it happen if you are not an expert in the field.
That's my 2cts at least.