Login
Register
Operational silos still slow fault resolution, complaint handling and network change. This Catalyst shows how agentic AI and digital twins can connect these domains in a governed, closed-loop framework for trusted AN L4 operations.
Agentic operations bring AN L4 closer to reality
As communications service providers move toward Autonomous Network Level 4, one of the biggest challenges is not simply automating more tasks. It is enabling operational domains to work together, reason across shared context and execute actions safely at scale.
Fault management, service assurance, customer complaint handling and change management are still often managed by separate systems and teams. That fragmentation slows issue resolution, increases operational cost and creates inconsistent customer experiences. Even where AI and automation are already in use, many solutions remain isolated and rule-based, limiting their ability to coordinate across customer, service and resource domains.
The Catalyst project, Driving agentic operations for ANL4, addresses this gap by creating a unified, closed-loop operational framework powered by agentic AI and Digital Twin Networks. Its goal is to give CSPs a practical and measurable path toward trusted AN L4 operations, where AI agents can sense, reason, decide and act across domains while maintaining governance, transparency and human oversight.
The Catalyst brings together fault management, complaint handling and change management into an adaptive operational ecosystem. Rather than treating each process as a separate workflow, the solution links them so that events in one domain can trigger, inform and govern action in another.
In practice, multi-agent AI can detect and analyze complex cross-domain anomalies across RAN, core, IP and transport networks. Customer complaints can be linked proactively to network and service degradation, helping teams understand business impact faster. Where infrastructure changes are required, the change management layer can be triggered automatically, with digital twins helping validate the likely impact before execution.
The framework is designed for progressive delegation. AI agents can handle low-risk, repetitive tasks autonomously, while high-impact or exceptional scenarios remain subject to human-in-the-loop control. This balance is critical for AN L4 adoption, because operators need automation that can act quickly without sacrificing safety, accountability or operational confidence.
At the center of the solution is an agentic AI framework integrated with retrieval-augmented generation, vector databases, digital twins and AIOps, LLMOps and AgenticOps capabilities. These elements give autonomous agents the operational knowledge, simulation capability and execution controls needed to move beyond alerting and recommendation into coordinated action.
One scenario focuses on agentic packet core fault management. AI agents correlate alarms, use RAG-powered root cause analysis and provide governed resolution recommendations in minutes. Another focuses on autonomous change management, where agents can plan, assess, validate and execute network changes through a policy-governed closed loop. Digital twins provide the assurance layer, giving teams a way to test impact and reduce risk before changes affect live network conditions.
The Catalyst draws on a broad set of TM Forum assets, including 10 Open APIs, eight Autonomous Networks assets, three DT4DI assets and eight AI Native Blueprint assets from the AI and Data mission. This standards-based foundation is important because AN L4 cannot scale through isolated vendor-specific automation alone. CSPs need reusable patterns that can connect operational domains, support explainability and enable trusted decision-making across complex environments.
The team expects the Catalyst to deliver significant operational and commercial benefits. In service assurance, it targets a 20% to 40% reduction in mean time to repair and 40% to 60% faster network changes. Across operational processes, the solution aims to automate 30% to 50% of activities, reducing manual intervention while improving consistency and speed.
The customer experience impact is also central to the business case. By connecting complaints with network and service context, the project targets 15% to 30% fewer complaints, together with improvements of 10% to 20% in first contact resolution and average handling time. For CSPs, that translates into lower cost to serve, faster restoration, higher service reliability and less customer churn.
Dr Aziza Najeeb Khamis Al Zadjali, Network Operations Strategist at Omantel, says the Catalyst “demonstrates how Agentic AI and Digital Twins can transform network operations, making it faster, smarter, and more precise, while accelerating progress toward AN L4 and beyond.”
For the wider industry, the Catalyst validates agentic AI and digital twins as key enablers of more autonomous, coordinated operations. For society, the benefits include more reliable digital connectivity, better service continuity, more efficient use of network and energy resources, and greater trust in AI-enabled critical infrastructure.
