This workshop will explore the evolution of Edge AI from traditional TinyML applications toward more capable, agentic systems that can reason, coordinate tools, and support real-world decision-making. Speakers will highlight how machine learning predictions can be combined with optimization, operational knowledge, human judgment, and AI agents to create practical decision systems rather than standalone models. Participants will also gain hands-on experience with small language models (SLMs), vision-language models (VLMs), and AI agents running directly on local edge devices. The session will emphasize both the growing technical capabilities of Edge AI and the importance of designing systems that deliver meaningful, measurable value.