This deliverable presents the design of the Decentralised AI-driven Swarmchestrate System and the specification of the four Demonstrator applications. The design targets to realise decentralised, self-organizing orchestration mechanism within the Swarmchestrate project which adopts a swarm-based, application-centric approach to resource management across the dynamic and heterogeneous Cloud-to-Edge continuum. Key achievements of the system design include: identification of the various subsystems and their roles in the Swarmchestrate system clarification of capabilities and functionalities of the different subsystems the Swarmchestrate system consists of designing the interfaces for each subsystems identification of the internal components for each subsystems specification of the deployment requirements for the components building up the different subsystems integration of the cooperating subsystems The project has 4 demonstrators for validating the decentralised, self-organising orchestration mechanism. For this purpose, a detailed analysis of the demonstrators and their requirements as well as the design of the demonstrator applications have taken place recently. Note that the video scene analytics platform (aimed to be deployed in the municipality of Trento by TRE and FBK) defined in the submitted proposal, has been replaced by a sound analytics demonstrator system (in the municipality of Koper, developed by InnoRenew, as of September 2024) due to legal issues, which change went through the official procedure via amendment, resulting in an accepted, updated grant agreement. Key achievements of the demonstrator specification include: understanding the operation of demonstrator applications elaborating the methodology on collecting technical and non-technical requirements of the demonstrator applications designing business and technical requirements questionnaire careful evaluation of the requirements questionnaire identification of the requirements of the demonstrators system design of the demonstrators identification of the metrics for monitoring and scaling performing the prototyping in its early stage The works introduced above forms the basis for implementing the decentralised, self-organizing orchestration and for implementing the four demonstrators in the second phase of the project.