Exposing agents as web services in JADE Orso Negroni 1 , Anthony Othmani 1 , Arthur Casals 2 , and Amal El Fallah Seghrouchni 1 1 Sorbonne Universit´ e, LIP6- CNRS UMR 7606 4 place Jussieu, 75005 Paris - France amal.elfallah@lip6.fr, orso.negroni@etu.upmc.fr, anthony.othmani@gmail.com 2 Escola Polit´ ecnica - Universidade de S˜ ao Paulo Av. Prof. Luciano Gualberto, 158 - trav. 3 - 05508-900 - S˜ ao Paulo - SP - Brazil arthur.casals@usp.br Abstract. The objective of this demonstration is to show how intelli- gent agents using a BDI architecture can be exposed as web services and integrated with existing cloud services. We will use JADE and BDI4JADE to expose an intelligent agent as a service that can be integrated with different cloud-based services, such as Amazon AWS services and Google Home. Keywords: Multi-Agents Systems · Web Service Agents · JADE/WSIG. 1 Introduction Using agents in conjunction with the Semantic Web is not a novel idea [3,7]. Additionally, model-driven development (MDD) techniques for developing agent models such as GAIA [9] and Tropos [1] can be used in the Semantic Web context [5] in order to allow the modelling of web-capable agents. Despite the extensive research already done in these fields, the advent of the Web of Things (WoT) [4] has encouraged a re-visitation on the use of agents in the context of the Semantic Web as a viable approach for deploying autonomous systems within the WoT [2]. Our research resides in the Agent-Oriented Software Engineering domain. During the study of existing agent modeling methodologies and their relationship with Web architectures, we decided to implement web-capable agents (exposed as web services) to serve as an evolving proof-of-concept, to be used during the course of our work. With that in mind, in this demo session we will present a Smart Agenda multiagent system (MAS), which is built to function as an agent-based personal assistant. For example, if the user updates his agenda with a meeting in Brussels and he is in Paris, the Smart Agenda system is supposed to detect that the user needs a train to go from Paris to Brussels, and use a text-to-speech service in or- der to ask Google Home to book a train ticket. We will detail our implementation in the next paragraphs.