It is believed that generative AI and agentic are two complementary concepts that are driving the automation and re-engineering of operations in any industry. Generative AI is the best at producing content, whether in the form of text, images, or code, in response to user inputs, whereas agentic systems take this ability and run everything independently and purposefully. This article discusses how the concept of reactively creating content began to change into proactively performing work. It characterizes both paradigms, discusses their technological enablers, such as LLMs, integration models, and orchestration of agents, and explains how agentic AI can be used to automate workflows, make decisions, and reinvent processes. We include case studies in different industries, analyse pros and cons, and write about architectural patterns like the agentic AI mesh item. We conclude with a description of challenging issues and future directions of research.