Introduction

In this tutorial, we will explore AI Agents and Small Language Models (SLMs) and their advanced implementation strategies for edge computing environments. We will cover the fundamental concepts of agentic AI, SLM optimization techniques, practical deployment strategies for resource-constrained devices, and Microsoft Agent Framework for building production-ready agent systems.

The landscape of artificial intelligence is experiencing a paradigmatic shift in 2025. While 2023 was the year of chatbots and 2024 saw a boom in copilots, 2025 belongs to AI agents — intelligent systems that think, reason, plan, use tools, and execute tasks with minimal human input, powered increasingly by efficient Small Language Models. Microsoft Agent Framework emerges as a leading solution for building these intelligent systems with offline edge-based capabilities.