2 ms·
As someone who recruits and interviews ML engineers, specifically for computer vision applications, I only barely look at your educational background. I look f
by AndrewKemendo 10y ago
As someone who recruits and interviews ML engineers, specifically for computer vision applications, I only barely look at your educational background.
I look for actual projects you have contributed to, published research, OSS contrib etc... that shows you can actually build something deployable and robust (mod your experience). Why? Most ML researchers are terrible at actually deploying products using ML.
School work can be relevant if it is part of a thesis, or research effort but in that case it's really still just [goto contributions].
At the application/implementation level you won't be making a new version of eg. gradient descent (and if you are you shouldn't be in industry as that's probably* a waste of resources), you'll be implementing existing ML systems and optimizing parameters. The most important thing you should be able to do is identify sources of data, structure data inputs flow and manage variability for the data you are using for both training and classification.
This doesn't answer your question directly, but it answers the implied question: What skills should I have to be a professional ML engineer?
- mempko 10y agoI can validate this. As someone who built a production CV system and interviewed graduates. Finding the magical ML graduate who can also code and make a production system is almost impossible.