In the last years a growing interest has arisen around a new paradigm called (among other names) Agents or Multi-Agents Systems. Its popularity has increased to the point that nowadays almost everyone working in Artificial Intelligence tries to present its job from the point of view of the Agent's paradigm. In this tutorial we try do clearly explain the reasons behind the success of this new paradigm and its advantages and disadvantages. We begin defining what an Agent and a Multi-Agents System are and what points have to be taken into account when designing them. We also explain different options that some authors have used to implement Agents and Multi-Agents System. We finish this work analysing some prototypical applications developed under the agent metaphor.