Distributed coordination represents managing inter-dependencies between positions and activities performed to achieve goals. Henry Mintzberg has defined the following mechanisms that can be applied to coordinate dependencies between positions and activities [1]: Direct Supervision (a central service takes all decisions by issuing instructions to agents and monitoring their actions), Standardization of Work (the specification of technical activity of the agent, step by step, without the possibility to negotiate, conflicts are reported to the supervisor), Standardization of output (agents coordinate by specifying the result), Standardization of skills (the specification of the competencies needed for the activity), and Mutual Adjustment (coordination by process of informal communication between agents). For distributed intelligent system design, such as Cloud-Edge Continuum [2], this means that agents have social abilities in the sense that they can interact and reason about each other’s interfaces, knowledge and competencies, and activities to achieve, without hardly any standardization or protocols. We proposed an architecture that has an essential layer for agent communication, management, and message transportation, as well as a specification for the abstract architecture and applications layers (see Figure 2).Figure 2: Proposed architecture using coordination and scheduling strategies