We explore different use cases in which machine-learning algorithms are used to handle non-trivial problems, involving large amounts of data items, and sometimes running in a low-resource environment (e.g., Raspberry Pi, Google Edge). After a brief introduction to machine learning and a few modern key algorithms, we will learn and experience some practical ideas, libraries and platforms, which will help us understand how machine learning is used nowadays in real-life applications. Our main coding language is Python; we will work with modern libraries and platforms, such as MLlib, PyTorch and XGBoost. We will survey machine learning topics, such as, decision trees, random forests, neural nets, topic modeling, recommender systems. Then we will delve into some of the most interesting problems for machine learning, like sentiment analysis, language modeling and text generation, word and text embeddings, object detection and more. Previous knowledge in machine learning is big plus, but attending the introductory machine learning course in parallel should suffice to enjoy this course.