High-fidelity data enabled AI agent’s swift resolution of a simulated issue while reducing data tokenization with AWS Bedrock by an estimated 80%
NETSCOUT SYSTEMS, INC., a leading provider of observability, AIOps, cybersecurity, and DDoS attack protection solutions, today announced its recent participation in TM Forum’s NeuroNOC Catalyst, an innovation project at DTW Ignite 2025, which set records for having the most Communication Service Providers and countries represented at a DTW Ignite event. TM Forum is a global alliance of telco and tech companies, leading the industry in defining the building blocks for new operating models, impactful new partnerships, and advanced software platforms. The NeuroNOC Catalyst project explored how AI agents, closed-loop automation, and high-quality network data can enable self-healing operations across telecom environments.
Developed in collaboration with global leaders, including Amazon Web Services, Accenture, Symphonica, and Sand Technologies—and championed by carriers such as BT Group, Telecom Argentina, Omantel, Turknet, Axian Telecom, and Safaricom—the NeuroNOC initiative demonstrated how AI and automation can dramatically accelerate fault detection and resolution in live network environments.
“While fully autonomous networks are still in their early stages, this project shows the powerful potential of AI agents – armed with the right data in real-time – to support faster and more accurate resolution of network issues.”
– Richard Fulwiler, Senior Director, Product Management at NETSCOUT
NETSCOUT deployed its Omnis AI Insights Solution, consisting of Omnis AI Sensor and Omnis AI Streamer, to deliver 5G Standalone Radio Access Network (SA RAN) and Packet Core high-fidelity telemetry essential for effective AI-driven operations. Omnis AI Sensor utilizes NETSCOUT’s signature deep packet inspection (DPI)-based, end-to-end network visibility, and Omnis AI Streamer provides powerful analytics and filtering at source via an open API driven dataset. In simulated service-impact scenarios, the solution enabled network operations center (NOC) engineers to identify subscriber registration issues, pinpoint the root cause via a curated large language model (LLM), and execute remediation steps with minimal manual effort.
The key takeaway was strong empirical validation of the need for high-quality curated data to drive effective AI solutions. Without data quality, AI models are hollow shells incapable of delivering valuable results, and data curation can unleash great power to ensure better outcomes. Moreover, the project showed promising results, including an expected 80% reduction in manual troubleshooting and up to 50% lower operational costs for communications service providers. It further reduced data usage and tokenization by AI models like AWS Bedrock by up to 80%, pointing to additional potential savings.
“Accurate, real-time curated data is the foundation of intelligent network operations. Without high-quality packet collection across the network, it’s nearly impossible to correlate issues across multiple data streams, determine root causes, and verify and test automated fixes,” said Richard Fulwiler, a Catalyst participant and Senior Director, Product Management at NETSCOUT. “While fully autonomous networks are still in their early stages, this project shows the powerful potential of AI agents – armed with the right data in real-time – to support faster and more accurate resolution of network issues.”
