8 insights from LogicMonitor’s Elevate event: AI-driven observability moves toward autonomous IT
SiliconANGLE Cheryl Knight
LogicMonitor says AI observability is pushing IT toward autonomous operations. GlobalFoundries cut alert noise 98%, which shows this is moving past hype.
Based on reporting by SiliconANGLE, Cheryl Knight — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
LogicMonitor used its Elevate event in Chicago to make a clear pitch: enterprise IT is no longer just about watching systems, but about using AI to act on what it sees. The company and its partners kept coming back to the same tension — more automation, but with more governance. That’s the balancing act now. Not fewer tools, exactly. Better control over the ones already in place.
The most eye-catching proof point came from GlobalFoundries. The chipmaker said Edwin AI cut alert noise by 98% and that its observability modernization is already 98% deployed after six months. In a factory setting, that matters because slow or noisy incident handling can turn into expensive downtime fast. GlobalFoundries is also consolidating observability tools, mapping dependencies in legacy applications and tying AI-driven insights into ServiceNow so issues can be handled before they disrupt production.
LogicMonitor framed this as a shift toward governed automation, not wild agentic everything. Karthik Sj and Garth Fort described a gradual path that starts with low-risk automation and then expands as trust builds. The company’s latest Edwin AI work adds multi-agent investigations, workflow integrations and more governance controls. OpenTelemetry and Catchpoint also sit inside that story, helping tie telemetry and application performance back to real operational decisions.
The customer examples were just as telling. Bell Techlogix is using AI assistants to speed up triage and documentation. ANS Group said Edwin AI improved engineering efficiency and customer retention, while Specialist Computer Centres described an AI agent that found an issue in about three minutes after engineers had spent 40 to 50 hours on it. That’s not a small tweak; that’s a different operating model.
And LogicMonitor is clearly trying to prove the same logic inside its own walls. Alyene Schneidewind said 95% of employees adopted ChatGPT within four months, and more than 2,000 custom GPTs were built in the first year. The company now says it is focused less on novelty and more on workflow redesign, business outcomes and the governance needed to keep autonomous IT from becoming autonomous chaos.
My take — AI-written commentary, not fact-checked reporting
This is the part of AI that actually matters: not the demo, but the control plane. Vendors can call it autonomous IT all they want; the real product is trust wrapped around automation, because nobody wants a fast system that confidently breaks things. The hype parade is over, and the boring governance bits are now the main event.
Read more about this at: SiliconANGLE