I've done two technical interviews this summer. The first one was shakey, the second went really well, but ultimately both places moved on with another candidate. It's definitely a skill and learning to not get so nervous about them helps a ton. I went from grad school physics to my first data science job and I'm trying to further transition into ML engineer type roles. My biggest insecurity is not knowing the comp sci 101 basics, since my programming was self taught along the way (learning what I had to learn for my thesis research simulations and then whatever else I learned in my first 2.5 years as a data scientist). I long for the day that I have enough experience under my belt that I'm a ML unicorn and companies view me as highly desirable, but I probably got many more years of grinding til I reach that point.
I got into software from a similar background. You can get pretty solid CS basics very quickly if you go through a book like Cracking the Coding Interview and do a few leetcode problems. It will also probably dramatically improve your experience writing code.
The best way to actually learn how to code is have someone else teach you whose been doing it for several years.
I wouldn't be where I'm at today had I not had my lead help me out during my internship. My school was just trash at teaching programming. Not only was it basically self taught, but everybody got A's so long as they just submitted code that compiled. My lead actually taught me how and why this stuff works.
