The critical role of Artificial Intelligence (AI) and Machine Learning (ML) in achieving highly automated management for next-generation mobile networks is undeniable. This paper elaborates on the DAEMON architectural model, which proposes a Network Intelligence (NI) Orchestration layer designed for effective end-to-end coordination of NI instances across the entire mobile network infrastructure. We first outline requirements and specifications for NI design, considering factors such as data management, control timescales, and network technology characteristics. Building on this analysis, we derive initial principles for designing the NI Orchestration layer. This includes proposals for the interaction loop between NI instances and the NI Orchestrator, as well as a unified representation of NI algorithms based on an extended MAPE-K model. The work ultimately contributes to defining the interfaces and operations of an NI Orchestration layer, aiming to foster the native integration of NI into future mobile network architectures.