For production machine learning / MLOps which is the core of this degree, there is an online course currently being rele

For technical interviews, I recommend the book Element of Programming Interview. The goal is not quantity but quality. Once you understand the common algorithms, most problems are just their variants.,To be more in-demand in the current job market by being end-to-end, and not subject yourself to the limits imposed by traditional educational systems, you need to be result-driven. Recognize your personal interests and the trend in technology, create our own path that leads to value creation.,A little digression before I dive into the structure and philosophy of this degree. I have an undergraduate degree in physics and used to study machine learning in grad school before deep learning became the hottest thing. I didn’t study computer science systematically but did take courses in algorithms and data structures. So you may see me fall into the “data science” camp mentioned above in terms of education. As for real-world software engineering and web development, I was clueless when I graduated.,But the story was not over. Closely following Go we got Rust. Initially I thought Rust was really what D should have been all along. A real rethinking of what C should have been. Rust kept low level control, high level abstraction mechanism and manual memory management but added unparalleled type safety. It all looked a bit too good to be true.,So if your goal is a career in building production-ready machine learning applications and not academic research or Kaggle competitions, the priority is to strengthen these areas and not to follow the next shiny model or tool every couple of weeks. This leads to the philosophy of how I structured my curriculum.,The road to a full-stack machine learning engineer may not look as glamorous as becoming a rockstar researcher. The current machine learning education online and offline often don’t even mention these factors that are more important than modeling in real-world ML applications:,General knowledge courses are for indexing knowledge in the brain into an organized, connected, and easy-to-search system. This is bottom-up learning and it is best when limited to the basics.,For machine learning and deep learning, I highly recommend fast.ai. It is the best top-down teaching on the internet and it’s free! Jeremy packs rich content in the lectures, you’ll have to watch the lectures more than once to get the most out of them. It all pays off when you build your own deep learning project using fastai techniques to achieve state-of-the-art results.,I went into the office every day feeling extremely insecure, grinding my way through each line of code, and somehow made it work without understanding the big picture how each component talked to each other.,Another way to group the courses is by knowledge domain. My degree has these major domains: computer science fundamentals, deep learning fundamentals, software engineering (including MLOps), and natural language processing.,It was initially painful but ultimately rewarding. I was lucky to have worked with an amazing team and learned a ton. During that period I also tried to make contributions to machine learning projects that were outside my team’s area. I began to realize the importance of software and data engineering. It is the most important aspect of any project, machine learning or not. Ask any machine learning professional, they will tell you that building models is just a tiny part of the job. Building something that people can actually use is almost an entirely different problem.,Project-based courses are where the real learning is. Come up with a small product you want to build, do some research on the necessary parts to learn, learn them on-demand as you build it up. When you are done, deploy it and try to get some users. Write an article teaching others by explaining the process in detail. It could become more useful than your resume in the future! This is top-down learning, an approach that is most effective in learning anything practical but is usually not adopted by traditional schools.,With a short period of intense preparation and some luck, I started my career as a software engineer at a Silicon Valley company, working on full-stack web development with some of the top engineers in the field.,After that, I joined an early-stage startup to experience what it’s like to be one of the founding members of a machine learning team. This is the period when I realized that formulating the right problem based on business use cases is the hardest of all tasks. The next hardest is how you collect and engineer your dataset. In all cases, modeling is a relatively minor factor of success, which might be very counterintuitive to beginners who came from classrooms where datasets and problems are well-defined and handed to them.,There is no need to understand every topic in-depth, but the key ideas are needed to be effective as an engineer. After going through these courses, at least you know what relevant info to look for when facing a new problem at work.