Practical MLOps

Chapter 7. MLOps for AWS

By Noah Gift

Everybody was scared of him [Dr. Abbott (because he yelled at everyone)]. When I was attending, there was a new resident named Harris. Harris was still afraid of him [Dr. Abbott] even though he was the chief resident and had been there for 5 years. Later [Dr. Abbott] has a heart attack and then his heart stops. A nurse yells, “Quick, he just had an arrest, come in!” So Harris went in there…leaning on the sternum and breaking ribs and stuff. So Harris starts pumping on Abbott and he woke up. He woke up! And he looked up at Harris and he said, “You! STOP THAT!” So Harris stopped. And that was the last story about Abbott.

Dr. Joseph Bogen

One of the most common questions I get from students is, “which cloud do I pick?” I tell them the safe choice is Amazon. It has the widest selection of technology and the largest market share. Once you master the AWS cloud, it is easier to master other cloud offerings since they also assume you might know AWS. This chapter covers the foundations of AWS for MLOps and explores practical MLOps patterns.

I have a long, rich history of working with AWS. At a sports social network I ran as the CTO and General Manager, AWS was a secret ingredient that allowed us to scale to millions of users worldwide. I also worked as an SME (subject matter expert) on the AWS Machine Learning Certification from scratch in the last several years. I have been recognized as an AWS ML Hero, I am also a part of the AWS Faculty Cloud ...