LogicMonitor is expanding its observability platform with more infrastructure data and safeguards for AI-driven IT responses, tying monitoring information to investigations while setting limits on automated actions.
At Elevate 2026 in Chicago, which ran Oct. 13-14, LogicMonitor outlined the changes on Oct. 14 across Envision, Edwin AI and Catchpoint insights. The company positioned the work under its Autonomous IT approach, which joins monitoring, AI-backed decision-making and actions subject to controls. The work covers five themes: enterprise data, linked service context, oversight of AI actions, user experience and trust.
The update brings in OpenTelemetry metrics, Windows Internet Information Services (IIS) application logs gathered by a Windows-based OpenTelemetry collector, and infrastructure cost data used in observability workflows. LogicMonitor said the extra feeds give investigations a broader set of operational clues.
The service context graph is intended to map relationships between infrastructure and applications, and the services and users tied to them. Topology-aware correlation can connect signals that may be related across that map, letting teams trace how a problem may affect other components.
Edwin AI investigations are expected to use purpose-built agents and workflows within LogicMonitor and tools such as Slack, Microsoft Teams and ServiceNow, with command-line interfaces also supported. A planned read-only MCP interface is intended to let compatible AI clients fetch LogicMonitor context, including Edwin AI information, with access controlled by authentication and permissions.
For AI-triggered operations, the company says policy checks, human sign-off, audit records and guardrails will limit automated execution. LogicMonitor gave no availability date for the additions.
LogicMonitor Chief Technology Officer Prabhu Nakkeeran said the transition would take time. “Autonomous IT is not a switch customers flip overnight. It’s a progression from trusted visibility, to better decisions, to governed action.”
