Mastering the Route
to Disaggregation

Leading in the development of open, disaggregated, virtualised and cloud native solutions, moving towards agentic AI-based operating models

“NGMN’s Route to Disaggregation defines how the mobile industry can move toward open, modular and interoperable network architectures in a controlled and sustainable way. By aligning operators and the wider ecosystem, the programme provides a clear roadmap for introducing disaggregated networks while preserving performance, security and operational stability.”

Johann Reindl

Senior Programme Manager (external), NGMN
johann.reindl@external.ngmn.org

NGMN Work Programme

„Cloud-native adoption provides the essential foundation for integrating advanced AI into telecom operations“.

Bernard Bureau

VP, Wireless Technology & Services, TELUS
NGMN Board Member

Cloud Native Next Steps

The project has been closed and the deliverable has been published in March 2026. Currently, the project team is identifying the next steps.
Shall you be interested in contributing please reach out to the Programme Manager Johann Reindl.

Network Automation

Project Lead

Project Co-Lead

Programme Manager

Objectives

Develop guidance for Agentic AI for Autonomous Mobile Networks by defining a mobile operator-centric reference framework.
Assess the relative priority and strategic importance of the selected scenarios by analysing where and why AI agents can address key operational challenges and enable business benefits.
Define a scenario-agnostic reference framework that specifies common components essential for Agentic AI collaboration, such as agent interaction patterns, lifecycle management, security models and interoperability interfaces.

Publications

Cloud-Native Next Chapter –
Agentic AI-Based Operating Models

March 2026

Building on NGMN’s Cloud Native Manifesto and established cloud-native frameworks such as the Cloud Native Computing Foundation’s (CNCF’s) Cloud Native Maturity Model (CNMM), the publication introduces a structured approach to integrate Agentic AI-based capabilities into telecom operations. It defines five progressive AI adoption levels and maps them to the CNCF Cloud Native Maturity Model (CNMM) stages, enabling operators to assess their readiness and required next steps to gradually evolve towards more intelligent and autonomous network operations. For each AI adoption level, the document also provides guidance on what is required across technology, people, skills and organisational culture.

MLOps for Highly Autonomous Networks

February 2025

This publication provides comprehensive guidelines and requirements specific to enabling technologies for highly autonomous networks. It focuses on a generalized ML (machine learning) model development and operation process, elaborating the MLOps (machine learning operations) process, requirements, architecture, and end-to-end deployment for Level 4+ autonomous network AI (artificial intelligence) applications. In this context, it examines the existing and relevant management standards applicable to AI applications and provides standardisation recommendations.

Operating Disaggregated Networks Operating Models – An Introduction

December 2024

A harmonised global 6G standard is essential to ensure a smooth industry evolution, address societal needs, and support a healthy ecosystem. The rollout of 6G should avoid mandatory hardware replacement, instead focusing on software upgrades within existing frequency bands. New hardware should only be introduced where new frequencies make it necessary. Any new technologies must demonstrate clear benefits within a techno-economic framework, with a strong emphasis on spectrum and energy efficiency.

Cloud Native Manifesto
An Operator View

September 2023

On our journey to highly flexible, sustainable, and resilient networks for the future, we believe in applying the following cloud-native principles to all layers of network infrastructure, applications, and services*:

Decoupled infrastructure and application lifecycles over vertical monoliths;
‘API-first’ over manual provisioning of network resources;
Declarative and intent-based automation over imperative workflows;
GitOps** principles over traditional network operations practices;
Unified Kubernetes (or the like) resource consumption patterns over domain-specific resource controllers;
Unified Kubernetes (or the like) closed-loop reconciliation patterns over vendor-specific element management practices; and
Interoperability through well-defined certification processes over vendor-specific optimisation.