In the past decade, the development of Artificial Intelligence (AI) and Edge Computing has impacted the way we analyse data in real time, make smart decisions, and build decentralized computing systems. This chapter discusses ongoing progress and difficulties with Edge intelligence used by healthcare, manufacturing, smart cities and autonomous systems by means of federated learning, lightweight neural networks and real-time analytics. It also takes on the biggest issues now, including thin resources, protecting privacy, optimising models and addressing security issues. Several experiments confirm that using edge-specific enhancements in deep learning and specialized hardware boosts performance and theoretical contributions give rise to new methods for federated learning, TinyML and on-device neural architecture search. These improvements do not completely solve the main problems with these models being used widely and at scale. Because of such findings, Edge AI is reshaping intelligence use by handling it in new and different places. Moving forward, we will need a computing fabric that is segmented, secure and efficient enough for the new kinds of applications we could not build before.