Understanding and solving fluid-related problems pose a significant demand for computationally inexpensive flow simulations, which the conventional fluid mechanics approach fails to satisfy. Thus, researchers choose to incorporate fluid mechanics with Machine Learning (ML) as a possible solution. This technology provides several tools and algorithms that help with prediction-based decision making, building optimized control theories, experience-based learning, and many more, all of which depend upon available data. Since its inception, the field of fluid mechanics has generated a lot in terms of experimental and simulation data. Hence, we can apply Machine Learning to extract meaningful information from fluid flow databases. Complex domains of fluid mechanics such as turbulence modeling, active flow control, and optimization all seek to gain from such a multidisciplinary approach. However, these domains from the world of fluids pose new problems for the data science world. These newly found complexities encourage engineers to create more robust learning models than conventional ones. Thus, a blend of fluid mechanics and machine learning creates a powerful and vastly complex field of study that will help completely revolutionize current research and industrial applications. This paper covers research from the earliest to some of the most recent ML algorithms and provides a brief overview of ways these algorithms complement the field of fluid mechanics. Three case studies—turbulence closure modeling using ML, flow control and manipulation using ML, and aerodynamic shape optimization, are used to explain this. To help better understand these applications, the underlying fundamentals of supervised, semi-supervised, and unsupervised learning models and some of the most widely used algorithms are also under consideration in the paper. The paper, thus, covers in great depth both the fields of fluids and ML. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.