This keynote will explore the rapid evolution of edge AI—from its early days running TinyML models to the emerging ability to deploy large language models (LLMs) and vision-language models (VLMs) directly at the edge. It will examine how this shift is transforming the IoT landscape by enabling real-time decision-making, reducing latency, and enhancing operational efficiency through on-device intelligence. Drawing from cutting-edge innovations in AI-enabled IoT modules, the session will highlight key trends such as the integration of AI accelerators across micro, thin, and thick edge layers, the convergence of edge computing with advanced IoT connectivity, and emerging use cases ranging from smart surveillance to autonomous robotics. The talk will also explore how edge AI reduces cloud dependency, drives cost-effective scalability, and enables decentralized innovation across sectors including automotive, industrial automation, and manufacturing.