DevOps for Machine Learning: MLOps

Interested in Machine Learning? Why not join my meetup group; MLOps Silicon Valley. It’s where Machine Learning scales!


MLOps (eh-m-eh-l-o-p-ss) is a term used to refer to the practices and techniques used to join the (mainly academic) world of machine learning theory to the needs of solving real-world machine learning problems.

Ever since the first computers were installed, there has always been a use for the “jack of all trades” type, who could do the kind of technical tasks that didn’t really fit with anyone else, but required a high degree of technical dexterity. “Web masters”, “Build Masters”, “DevOps”, “Release Engineer”, “Full Stack Developers”, “Operations Engineers”, to name but a few.

Ask any of these people what their job actually entails, and you’re likely to see them flounder. Why? Because, these are the sort of people who automate any task they do more than once. If they could tell you what they do every day, then they’re probably not doing it right.

Now in the world of machine learning, we have “MLOps”. As machine learning hits the corporate world, we’re going to need a lot of these types of people. Train them up. Buy them pizza. Keep them close. It’s hard to define what it is you’ll need from for, but one thing is for certain; if you don’t find them now, this is where your skills gap will be.