This talk presents the ACES Knowledge Graph (KG) model, and its role at fostering autopoiesis within cloud-edge infrastructures. Autopoiesis in this context refers to the system’s ability to self-manage, sustaining its own operations through dynamic adaptation. The ACES Knowledge Graph Model structures complex relationships across distributed cloud and edge systems, capturing interactions, dependencies, and temporal changes in a consistent, standards-based format. This temporal dimension of the model is crucial, allowing the KG to capture temporal dependencies and interactions, as well as update dynamically to facilitate reasoning processes as the system evolves. The model’s runtime capabilities support real-time adaptation to operational demands and anomalies, with mechanisms to detect behavioral patterns through temporal windows and motifs. We illustrate how the KG enables autopoietic behavior by supporting reasoning components that can predict demand/supply across the infrastructure, detect anomalies, and identify potential errors and attacks. This enables the swarm intelligence and scheduling optimization techniques to adapt the whole system across layers of the edge cloud continuum, as the KG captures the relationships across metrics, components and entities from all the infrastructure.