Role-Based Agentic AI for Intent-Driven Network and Service Orchestration

Researchers propose a role-based multi-agent AI system for telecommunications networks that bridges business and operational support systems through intent-driven orchestration. The framework applies hierarchical agent coordination to automate complex network management while maintaining privacy and accountability across organizational domains.

This research addresses a critical gap in telecommunications infrastructure management where existing AI solutions operate in silos between business-level strategy and operational execution. The proposed multi-agent system architecture mirrors real CSP organizational structures, creating a natural alignment between technical implementation and business accountability. This is significant because telecommunications networks represent some of the world's most complex distributed systems, and autonomous management through agentic AI could substantially reduce operational costs and improve service quality.

The hierarchical four-layer approach—spanning customer engagement, strategic planning, service delivery, and infrastructure provisioning—represents a meaningful evolution beyond previous intent-based networking attempts. By explicitly decomposing functional responsibilities across specialized agents, the framework enables organizations to maintain domain expertise while achieving coordinated end-to-end service orchestration. This mirrors successful patterns in enterprise software architecture and distributed systems design.

For telecommunications operators and infrastructure providers, this research suggests a viable path toward autonomous network management that preserves existing organizational structures rather than requiring wholesale transformation. The privacy-preserving domain separation is particularly relevant given regulatory requirements across different regions and the sensitive nature of customer data. As 5G and 6G deployments increase network complexity exponentially, autonomous management systems like this become increasingly important for competitive differentiation and operational efficiency.

• →Multi-agent AI architecture successfully bridges business and operational support systems in telecommunications networks
• →Hierarchical agent design mirrors CSP organizational structures, improving accountability and scalability
• →Privacy-preserving domain separation enables secure cross-functional coordination without compromising data governance
• →Proof-of-concept demonstrates feasibility of intent-driven orchestration from customer requirements to infrastructure provisioning
• →Framework addresses growing complexity in heterogeneous networks where autonomous management becomes economically necessary