Facebook
Britain's News Portal
Around The Clock
BREAKING
Loading latest headlines…

NETSCOUT: AI Operations Need Forensic-Grade Network Data

A sponsored feature from NETSCOUT argues that AI-driven operations require high-fidelity network data beyond traditional metrics, events, logs and traces, as observability gaps risk scaling uncertainty in autonomous systems.

  • NETSCOUT research cited in the feature says 81 percent of organisations believe insufficient data increases incident resolution time, and 42 percent estimate downtime at $500,000 to $999,000 per hour.
  • The feature says 96 percent of organisations use metrics and logs, yet 82 percent report visibility gaps and 96 percent lack sufficient data to determine root cause during incidents.
  • NETSCOUT director Jack Callahan says feeding partial, periodic or sampled data to an AI agent risks scaling uncertainty quickly.

Organisations adopting AI-driven operations need a data foundation that can support autonomous decision-making at machine speed, according to a sponsored feature from NETSCOUT published by The Register. The feature argues that traditional observability built on metrics, events, logs and traces (MELT) cannot by itself keep pace with the complexity and scale of modern digital infrastructure.

Jack Callahan, director of enterprise strategy at NETSCOUT, said executives often find the data available to them is not as conclusive as they would want. "And therefore, they're trusting their gut more than they'd expect, given how much they're spending," he said.

The feature cites NETSCOUT research indicating that 81 percent of organisations believe insufficient data increases incident resolution time, while 42 percent estimate downtime at $500,000 to $999,000 per hour. It also says 96 percent of organisations use metrics and logs, yet 82 percent report visibility gaps and 96 percent lack sufficient data to determine root cause during incidents.

According to the feature, these gaps become more serious when systems operate autonomously. Callahan said an AI agent will make decisions based on the data it has, so feeding it partial, periodic or sampled data risks scaling uncertainty quickly. The feature says only 41 percent of organisations describe AI-assisted insights as very or extremely consistent, 38 percent admit lacking forensic-grade data to validate automated actions, and 28 percent do not fully trust automation output.

NETSCOUT proposes enriching MELT data with packet-derived metadata, an approach it calls MELT+, to provide transaction-level evidence of what actually traversed the network. The feature says this is intended to complement rather than replace existing observability investments.

Why this matters: The feature argues that as enterprises adopt AI-driven operations, incomplete or sampled telemetry could lead to false correlation, ambiguity over root cause and overconfidence in partial signals.

Related Articles

Get the news that matters.

Join thousands of readers getting the best of British news straight to their inbox.