20 books and 1 piece of advise for aspiring ML / AI Engineers
The books that helped me stop feeling lost (most of the time)
One of the questions I get asked all the time is:
What resources do you recommend to get started in ML / AI Engineering?
Sounds like a simple enough question, right? It’s not.
Sure, I could drop a giant list of courses, books, articles, videos, blog posts … the whole buffet.
You’d open 200 tabs, feel hyped for five minutes, and then slowly spiral into a tab-induced meltdown. I call it the Wikipedia Effect, and yes, I’ve been there more times than I’d like to admit 😅
So instead of overwhelming you with everything, I’m going to keep it focused.
I’m talking:
20 books (yes, books still matter),
1 piece of advice (not a resource, but possibly the most important thing when things get tough)
Ready? Let’s go!
📕 The 20 books
What you’re looking at is my bookshelf.
(Well, the technical one. There’s another one for Physics, another for Philosophy … maybe one day I’ll give you the full tour 😅)
These are the books that have stuck with me the most: my favorites.
Sure, I’ve got others scattered around the house, but this shelf? This one’s curated. Okay, minus the one I accidentally left at my parents’ place months ago. Still counts.
Unlike the rest of my room (which thrives in chaos), this shelf has structure.
I’ve grouped the books into four categories that I think cover pretty much everything you need to become a solid ML / AI Engineer.
Let’s start with group one.
1️⃣ The Fundamentals
Book 1 - Mathematics For Machine Learning
One of the biggest mistakes I see beginners make in ML / AI engineering?
Skipping the fundamentals.
And no, I’m not saying you need to become a wizard in Hilbert Spaces or invent your own optimization algorithm.
I’m also not saying you need to read every page of the four books I’m recommending (though I do suggest going cover-to-cover on Mathematics for Machine Learning).
But the best ML/AI engineers I’ve worked with?
They get the math.
Not at a PhD level, but enough to understand what’s under the hood. They can think beyond LLMs (because no, not everything is an LLM). They can sit in a room with a Senior Data Scientist and have a real conversation.
I mean, if our job is to optimize, deploy, monitor, and improve ML models …
How are we supposed to do that without solid fundamentals?
2️⃣ Hands-on Learning
Book 5 - Machine Learning with Pytorch and Scikit-Learn
Book 6 - Hands-on Machine Learning with Scikit-Learn, Keras, and Tensorflow
Book 7 - Generative Deep Learning
Now that the basics are covered, it’s time to introduce you to some hands-on learning gems.
Whether you’re building confidence with PyTorch and Scikit-learn, exploring the different algorithms behind generative deep learning, or (like me lately) getting obsessed with deep reinforcement learning … there’s a lot to play with.
If you’re serious about understanding LLMs, I highly recommend the brilliant book Build a Large Language Model (From Scratch) by
.3️⃣ Data Engineering
Book 10 - Designing Data-Intensive Applications
Book 11, 12 - Learning Spark (1st Edition and 2nd Edition)
You might be wondering:
What’s this guy’s problem with Spark? Does he have a weird obsession or something?
Not at all. In fact, I think knowing Spark is one of the most underrated ways to stand out as an ML Engineer.
You only start to truly appreciate distributed computing when you're dealing with billions of records. That’s when Spark stops being “just another tool” and starts becoming your best friend.
So, my advice?
If you want to be the kind of ML Engineer who’s not scared of real Big Data, get your hands on a book like Learning Spark. It’s a solid starting point.
And while we’re on the topic:
If you haven’t read Designing Data-Intensive Applications, you’re missing out. That book is essential for anyone who’s serious about building solid, scalable systems. It’s one of those reads that separates decent engineers from great ones.
Amazon, TikTok, Spotify, … they didn’t get where they are just because of fancy algorithms. They got there because of world-class Data Engineers making the data flow like clockwork so those models can actually work.
So yeah, don’t sleep on Data Engineering. Ever 🧐
4️⃣ ML / AI Engineering
Book 15 - Scaling Machine Learning with Spark
Book 16 - Generative AI on Google Cloud with LangChain
Book 17 - Designing Machine Learning Systems
Finally! We’ve arrived at the good stuff.
This part of my shelf is all about ML Engineering and AI Engineering.
But, before we begin, it helps to actually understand what those roles mean. Thankfully, there are two books that do a fantastic job breaking that down: Designing Machine Learning Systems and AI Engineering, by
.These two books helped me build a solid conceptual map of both fields: what they are, how they overlap, and where they diverge.
Now, if you’re looking for something more applied and closer to real-world applications, I’d recommend going for books that focus on deployment, scalability, and monitoring.
Let me point you to one of my personal favorites:
The LLM Engineer’s Handbook, by
and .💬 The advise
After this little tour of my shelf, it’s time for that piece of advice I promised you.
I know it’s not exactly a “resource,” but maybe it’ll be a bit of light when “the night is dark and full of terrors”.
And no, I’m not talking about White Walkers, I’m talking about the everyday monsters: analysis paralysis, FOMO, the constant flood of information that makes you feel like you're always behind.
I’ve been there.
But you need to understand something:
You’re not going to become a Senior ML / AI Engineer in a month. Not in a year, either. Honestly, what even is a Senior AI Engineer? This field is still so young … 🤣
The key?
Take it one step at a time. Enjoy the ride. Because if you don’t love what you’re doing, this field can quickly turn into a nightmare. It’s demanding, intense, and never-ending if you don’t pace yourself.
Don’t skip the fundamentals just because the latest agentic framework looks shiny and cool. Sure, explore it—but build that solid foundation first. Your future self will thank you for it.
And let’s be real: it’s not just about books. Yeah, I’m aware this article is about books (guilty), and I do love reading. But I’ve made the mistake of thinking that just reading 20 books automatically made me a great ML engineer. Well, it didn’t.
The real growth comes when you take that knowledge and build something with it. Apply what you learn. Get your hands dirty. That’s how you level up.
So, my advice?
BUILD. BUILD. BUILD.
About how long in your estimation would it take to get through this set of books? (I have a full time SWE job)