NETSCOUT expands data platform to support AI-driven network operations

NETSCOUT Systems, Inc. is expanding its data platform to provide a layer of operational evidence designed to support enterprise AI, observability, cybersecurity, and service assurance. The platform aims to convert network packets into compact, contextualized evidence in real time, addressing the issue of fragmented or noisy telemetry that often hinders AI decision-making.

By providing high-fidelity data at the point of observation, the company seeks to reduce the need for AI systems to reconstruct events from sampled or aggregated metrics, events, logs, and traces (MELT). This approach, described by Chief Operating Officer Sanjay Munshi as ‘context engineering’, aims to improve the accuracy of AI recommendations while lowering compute and inference costs.

Internal testing by NETSCOUT reportedly showed a reduction in AI token consumption of more than 25% compared to using MELT data alone. Furthermore, the company noted a reduction in Mean Time to Knowledge (MTTK) of more than 75%. These developments align with Gartner research suggesting that prioritizing semantics in AI-ready data could improve agentic AI accuracy by up to 80% and reduce costs by up to 60% by 2027.