Pragmatic AI: An Introduction to Cloud-Based Machine Learning

by Noah Gift

Master Powerful Off-the-Shelf Business Solutions for AI and Machine Learning
Pragmatic AI will help you solve real-world problems with contemporary machine learning, artificial intelligence, and cloud computing tools. Noah Gift demystifies all the concepts and tools you need to get results--even if you don't have a strong background in math or data science. Gift illuminates powerful off-the-shelf cloud offerings from Amazon, Google, and Microsoft, and demonstrates proven techniques using the Python data science ecosystem. His workflows and examples help you streamline and simplify every step, from deployment to production, and build exceptionally scalable solutions. As you learn how machine language (ML) solutions work, you'll gain a more intuitive understanding of what you can achieve with them and how to maximize their value. Building on these fundamentals, you'll walk step-by-step through building cloud-based AI/ML applications to address realistic issues in sports marketing, project management, product pricing, real estate, and beyond. Whether you're a business professional, decision-maker, student, or programmer, Gift's expert guidance and wide-ranging case studies will prepare you to solve data science problems in virtually any environment.
Get and configure all the tools you'll need Quickly review all the Python you need to start building machine learning applications Master the AI and ML toolchain and project lifecycle Work with Python data science tools such as IPython, Pandas, Numpy, Juypter Notebook, and Sklearn Incorporate a pragmatic feedback loop that continually improves the efficiency of your workflows and systems Develop cloud AI solutions with Google Cloud Platform, including TPU, Colaboratory, and Datalab services Define Amazon Web Services cloud AI workflows, including spot instances, code pipelines, boto, and more Work with Microsoft Azure AI APIs Walk through building six real-world AI applications, from start to finish Register your book for convenient access to downloads, updates, and/or corrections as they become available. See inside book for details.

272 pages, Kindle Edition

Published July 12, 2018

About the author

Noah Gift is lecturer and consultant at UC Davis Graduate School of Management in the MSBA program, Northwestern's Master of Data Science program, and UC Berkeley Data Science program. He is also the founder of Pragmatic AI Labs. At Pragmatic AI Labs he provides consulting and training on Machine Learning and AI, and also develops AI SaaS products. At UC Davis, he is teaching graduate machine learning and consulting on Machine Learning and Cloud Architecture for students and faculty. He has published close to 100 technical publications including two books on subjects ranging from Pragmatic AI: An Introduction to Cloud-Based Machine Learning, First Edition to DevOps. He is also a certified AWS Solutions Architect and has an MBA from UC Davis, a M.S. in Computer Information Systems from Cal State Los Angeles, and a B.S. in Nutritional Science from Cal Poly San Luis Obispo. Professionally, Noah has approximately 20 years’ experience programming in Python and is a member of the Python Software Foundation. He has worked in roles ranging from CTO, General Manager, Consulting CTO and Cloud Architect. This experience has been with a wide variety of companies including ABC, Caltech, Sony Imageworks, Disney Feature Animation, Weta Digital, AT&T, Turner Studios and Linden Lab. In the last ten years, he has been responsible for shipping many new products at multiple companies that generated millions of dollars of revenue and had global scale. Currently he is consulting startups and other companies, on Machine Learning, Cloud Architecture and CTO level consulting via Pragmatic AI Labs

Reviews

[Emre Sevinç · Rating 3 out of 5]

If possible, I'd give this one 2.5 stars and I'm sad to say that because the author seems to be a very senior, experienced software engineer with many interesting projects under his belt, and Pearson Addison-Wesley is a respected publisher, therefore I don't understand why there wasn't enough editorial effort and it seems so rushed, and some sections so half-baked.

One of the serious problems seems to be catering to an audience: Very junior people will be much better served by other high quality material for Python skills (20 pages of "pragmatic" introduction barely scratches the surface). On the other hand, if someone has already programmed in Python for a few years (either in a research, or a professional setting), even without any AI or data science work, then such an introduction is wasted number of pages for them (that would be put to good use for more in-depth analysis of other topics). Of course, similar criticism applies to author's handling of popular Python data science, ML and visualization libraries such as pandas, scikit-learn, and others.

I get that the author's focus is on being "pragmatic", in the sense to quickly show what's available, and how interesting technology offerings can be put to concrete use to provide value, but even that doesn't justify skimming only the bare surface of some cloud offerings.

I don't want to be harsh; as a senior IT professional who had his fair share of training people and creating material to be read I know how difficult it is. The author's repeated focus on setting up local development environment for Python-based ML, AI and cloud development deserves respect. Also his use of Makefile based workflow automation is a lesson for younger generation of developers. I'd even add that this topic could easily deserve a separate, mini chapter on its own! On top of that, the author also is keen of always supporting his AI/ML development workflow by creating small and self-contained CLI (Command Line Interface) applications / utilities, and indeed, this is a very good practice; having such a "toolsmith" mentality will help many developers become more efficient, because it acts like productivity multiplier.

I also found a few chapters under the "III. Creating Practical AI Applications from Scratch" section well-balanced with useful demonstrations and nuggets of information. Those chapters can be valuable to demonstrate to junior software developers and starting researchers the flexibility of Python and its libraries, as well as visualization capabilities.

On top of that, I think dedicating a chapter on optimizing AWS instances, how to analyze them, how to manage costs (e.g. by relying on Spot instances and avoiding risks for unexpected costs) and strategies for automatically optimizing for cost is a good idea. Of course, this topic can easily lead to a mini-book by itself, but creating awareness for all software engineers regarding the cost of cloud services is important.

As I said in the beginning, I found this book pretty rushed, lacking high quality editorial support. For new starters some parts will be overwhelming and lacking details, though they might be enough to provoke curiosity. For experienced people it'll be very hit and miss, and many pages will seem like waste.

PS: I wonder the reasoning behind using the same quote in the beginning of two chapters (7 and 11), maybe another indication of lack of proper editorial process?

[Ronald Del · Rating 5 out of 5]

Good intro

Get your hands dirty quick on how to run machine learning workloads in the cloud. Learn modern developer tooling and workflows to help you on the job

[Jay · Rating 3 out of 5]

I think Noah Gift is probably a genius, and I think he may also be a bull in a china shop. I somewhat doubt that if he’d been alive in the 40’s that humanity would have survived the atomic age. As I’m reading Daniel Kahneman right now, I find myself longing for slow thinking in the midst of Mr. Gift’s glorious display of expert intuition.

To be fair, the book is more what it promises to be and less what I expected to be.

What I expected and hoped for was a guide to making decisions about AI tools and making smart choices about building an AI toolchain. What I got was blisteringly paced look into what tools Mr. Gift has chosen and fine grained examples of how he has used them.

For my purpose, I’m able to infer some lessons about the tools by paying attention to the praise and criticism he offers to various tools, the characteristics of those tools he selects as a focus, and the themes he comes back to multiple times in his cookbook examples.

Coming away from a quick read of the book, I’m struck by a few strong impressions. One is the perspective Mr. Gift brings on pragmatism in software engineering, what in other venues might be described as “the good versus the perfect.” I’m also left astounded at both the mastery the author demonstrates over a large spectrum of highly complex instruments of analysis and his ability to demonstrate that mastery in such a short work. It is a joy to see, and worth the price of admission for someone who studies education.

I’m also left with an impression of frenzy and am left wondering about the interaction between frenzy and pragmatic choices. I think I should be forgiven the feeling of mental whiplash as Gift bounds from Amazon to Google to R, always coming back to his base of Python.

So, all of that is to say that I’m fascinated by the book, but I can’t decide if I’m educated by it. Time will tell how the exposure to Pragmatic AI will influence my decision making as an analytics leader going forward.

[Oge · Rating 2 out of 5]

This book doesn't do justice to the author's knowledge and experience in the fields of AI and ML.
You are better off going to the companion github page (https://github.com/noahgift/pragmaticai) and checking out Jupyter notebooks, video lectures, and other resources available there.

[Abhijith Nayak · Rating 2 out of 5]

The editing, in the Indian black and white edition, is terrible. You'll still be able to gather a few pointers but it seems like no one even bothered to proofread the book, and the fonts and typography are a strain on the eye.

[Horia Calborean · Rating 3 out of 5]

This book was not for me. If you are already good at ML then this book might give you some hints.

[Loc Nguyen · Rating 5 out of 5]

Skimmed the book - and it feels less like an introduction than it does a reference book.

[Swamy Atul · Rating 2 out of 5]

I skimmed the Python codes. The most useful sections were the one where the practitioner wisdom of the author is presented in a no-nonsense blink-and-miss sentences.