Innovating new M-commerce applications have been slow in coming to market and deployments so far have been primarily limited to simple infotainment services such as downloading screen savers and ring tones, or getting the latest news. While some of the reasons for this are market related there are also a number of very challenging technical problems that underlie the deployment of true end-to-end M-commerce user experiences. Chief amongst these are: 4. Coherent, consistent and reliable capture of user context in terms of location, service offerings, other users, environmental conditions. 5. The management of trust and privacy services/information ensuring users data is protected yet can be leveraged to the user's benefit where appropriate. 6. Effective management and control of user service discovery, access, composition and hence service semantics to achieve complex goals requiring multiple service interactions. Success in each of these areas is critical to the provision of seamless, consistent, ubiquitous service offerings. Furthermore solutions require developments at many levels of the network from the' software in the users handset to services deployed in the network While there are a considerable number of promising technologies and initiatives which may help address these challenges such as agent technologies for mobile devices, agent negotiation, Markup languages for knowledge representation, Web service semantics, Web service registries, Web standards related with security and privacy; it is unclear how they can be adequately combined, deployed and used in a real production environment. The Telesto project aims to bring together industry and research organizations to develop a vertically integrated, production-ready demonstration environment that addresses the problems outlined above. In particular we believe that a key role will be played by: agent technology to automate the chores that previously required human attention, semantic web technology to make Web services understandable to agents and web services to automate Web service discovery, composition and execution through intelligent agents. The project will involve: core and applied research for adaptation of existing formalisms and the development of new ones, development and deployment to create the demonstration environment, a significant field trial to test the resulting system, take up provisions to enable the transfer of results into real usage.