Chapters

The first few chapters cover the theory and practice of both DevOps and MLOps. One of the items covered is how to set up continuous integration and continuous delivery. Another critical topic is Kaizen, i.e., the idea of continuous improvement in everything.

There are three chapters on cloud computing that cover AWS, Azure, and GCP. Alfredo, a developer advocate for Microsoft, is an ideal source of knowledge for MLOps on the Azure platform. Likewise, Noah has spent years getting students trained on cloud computing and working with the education arms of Google, AWS, and Azure. These chapters are an excellent way to get familiar with cloud-based MLOps.

Other chapters cover critical technical areas of MLOps, including AutoML, containers, edge computing, and model portability. These topics encompass many cutting-edge emerging technologies with active traction.

Finally, in the last chapter, Noah covers a real-world case study of his time at a social media startup and the challenges they faced doing MLOps.

Appendixes

The appendixes are a collection of essays, ideas, and valuable items that cropped up in years between finishing Python for DevOps (O’Reilly) and this book. The primary way to use them is to help you make decisions about the future.

Exercise Questions

In this book’s exercises, a helpful heuristic considers how you can leverage them into a portfolio using GitHub and a YouTube walkthrough of what you did. In keeping with the expression “a picture is worth a thousand words,” a YouTube link to a walkthrough of a reproducible GitHub project on a resume may be worth 10,000 words and puts the resume in a new category of qualification for a job.

As you go through the book and exercises, consider the following critical thinking framework.

Discussion Questions

According to Jonathan Haber in Critical Thinking (MIT Press Essential Knowledge series) and the nonprofit Foundation for Critical Thinking, discussion questions are essential critical thinking components. The world is in dire need of critical thinking due to the proliferation of misinformation and shallow content in social media. A mastery of the following skills sets an individual apart from the pack:

Intellectual humility

Recognition of the limits of your knowledge.

Intellectual courage

The ability to argue for your beliefs even in the face of social pressure.

Intellectual empathy

The ability to put yourself in the minds of others to understand their position.

Intellectual autonomy

The ability to think for yourself independently of others.

Intellectual integrity

The ability to think and argue with the same intellectual standards you expect others to apply to you.

Intellectual perseverance

The ability to provide evidence that supports your position.

Confidence in reason

The belief that there are indisputable facts and that reason is the best solution to gain knowledge.

Fairmindedness

The ability to put in the good-faith effort to treat all viewpoints fairly.

Using these criteria, evaluate the discussion questions in each chapter.

Origin of Chapter Quotes

By Noah

I graduated college in late 1998 and spent a year training to play professional basketball in the minor leagues in the United States or Europe while working as a personal trainer. My backup plan was to get a job in IT. I applied to be a Systems Administrator at Caltech in Pasadena and got a Mac IT expert position on a fluke. I decided the risk/reward ratio of being a low-paid professional athlete wasn’t worth it and accepted the job offer.

To say Caltech changed my life is an understatement. At lunch, I played ultimate frisbee and heard about the Python programming language, which I learned so I would “fit in” with my ultimate frisbee friends, who were staff or students at Caltech. Later, I worked directly for Caltech’s administration and was the personal Mac expert at Caltech for Dr. David Baltimore, who got the Nobel Prize in his 30s. I interacted with many famous people in many unexpected ways, which boosted my self-confidence and grew my network.

I also had many Forrest Gump–style random encounters with people who would later do incredible things in AI/ML. Once, I had dinner with Dr. Fei-Fei Li, head of AI at Stanford, and her boyfriend; I remember being impressed that her boyfriend spent the summer writing a video game with his dad. I was highly impressed and thought, “Who does that kind of thing?” Later, I set up a mail server under the famous physicist Dr. David Goodstein’s desk because he kept getting grief from IT about hitting his mailbox storage limits. These experiences are where I acquired a taste for building “shadow infrastructure.” Because I worked directly for the administration, I got to flaunt the rules if there was a good reason for it.

One of the people I randomly met was Dr. Joseph Bogen, a neurosurgeon and visiting professor at Caltech. Of all the people I met at Caltech, he had the most profound impact on my life. One day I responded to a help desk call to come to his house to fix his computer, and later this turned into a weekly dinner at his home with him and his wife, Glenda. From around 2000 until the day he died, he was a friend and mentor.

At the time, I was very interested in artificial intelligence, and I remember a Caltech Computer Science professor telling me it was a dead field and I shouldn’t focus on it. Despite that advice, I came up with a plan to be fluent in many software programming languages by 40 years old and writing artificial intelligence programs by then. Lo and behold, my plan worked out.

I can clearly say I wouldn’t be doing what I am doing today if I didn’t meet Joe Bogen. He blew my mind when he told me he did the first hemispherectomy, removing half of a brain, to help a patient with severe epilepsy. We would talk for hours about the origins of consciousness, the use of neural networks in the 1970s to figure out who would be an Air Force pilot, and whether your brain contained “two of you,” one in each hemisphere. Above all, what Bogen gave me was a sense of confidence in my intellect. I had severe doubts up until that point about what I could do, but our conversations were like a master’s degree in higher-level thinking. As a professor myself, I think about how big of an impact he had on my life, and I hope to pay it forward to other students I interact with, both as a formal teacher or someone they meet. You can read these quotes yourself from an archive of Dr. Bogen’s Caltech home page and his biography.

From Noah

As mentioned earlier, without Mike Loukides inviting me to Foo Camp and having a great discussion with Tim O’Reilly and me, this book wouldn’t be here. Next, I would like to acknowledge Alfredo, my coauthor. I have had the pleasure of writing five books, two for O’Reilly and three self-published, in a little over two years with Alfredo, and this is primarily due to his ability to embrace work and get things done. An appetite for hard work is perhaps the best talent, and Alfredo has this skill in abundance.

Our editor, Melissa Potter, did tremendous work getting things into shape, and the book before she edited and afterward are almost two different books. I feel lucky to have worked with such a talented editor.

Our technical editors, including Steve Depp, Nivas Durairaj, and Shubham Saboo, played a crucial role in giving us great feedback about where to zig and when to zag. Many enhancements are particularly due to Steve’s thorough feedback. Also, I wanted to thank Julien Simon and Piero Molino for enhancing our book with real-world thoughts on MLOps.

I want to thank my family, Liam, Leah, and Theodore, for giving me the space to finish this book on a tight deadline in the middle of a pandemic. I am also looking forward to reading some of the books they write in the future. Another big group of thanks goes out to all the former students I taught at Northwestern, Duke, UC Davis, and other schools. Many of their questions and feedback made it into this book.

My final thanks go out to Dr. Joseph Bogen, an early pioneer in AI/ML and Neuroscience. If we didn’t bump into each other at Caltech, there is zero chance I would be a professor or that this book would exist. His impact was that big on my life.