ScienceLogic, delivering trusted operational intelligence through service-centric observability, AI-driven operations, and intelligent automation, today announced Skylar AI 2.8, introducing new agentic AI specialists that investigate IT issues and help teams get to the root of problems faster. IT teams often spend valuable time determining whether an alert represents a real problem, investigating its cause, and deciding what to do next. The new Topic Specialist Agents in Skylar Advisor™ take on that investigative work, using live telemetry to deliver evidence-backed findings, recommendations, and approval-gated remediation plans. Administrators control specialist enablement and scheduling, giving IT teams a governed way to investigate issues and assess next steps while maintaining oversight and control.
Skylar AI 2.8 represents another step toward agentic IT operations, moving beyond AI-generated recommendations to agents that actively investigate issues. With evidence-backed findings and human oversight built in, organizations can put AI to work while maintaining the visibility and control they need.
Additional enhancements in Skylar AI 2.8 give organizations new ways to investigate advisories, work through complex operational questions, and coordinate team response. This release also strengthens security controls and further enhances OpenTelemetry metric ingestion and alerting. Key functionality includes:
- Topic Specialist Agents: Specialist agents investigate advisories using live telemetry to assess whether an issue is genuine, explore likely causes, and recommend next steps. Each specialist brings a focused investigation approach to its topic, such as network, storage, or database issues, and provides cited evidence and confidence scores so teams can evaluate the findings.
- OpenTelemetry metrics and Alert Builder (preview): When enabled, organizations can stream OpenTelemetry gauge and sum metrics into Skylar AI through a provided collector agent and build alerts on those metrics.
- Multistep investigation in chat: Multistep answers let Skylar Advisor investigate complex questions through plans in which later steps build on earlier findings. The release also improves follow-up handling and automatic tool selection. New reporting tools rank devices by resource use, forecast capacity, and report on device inventory and membership.
- Role-based team workflows: A rebuilt Recent Activity dashboard gives operators, supervisors, and business owners views tailored to their roles. Supervisors can assign work, review proposed assignments, and coach team members. Health scoring covers services, devices, and organizations. An automatic sweep closes idle advisory threads using a configurable threshold so stale work does not distort workload and health metrics.
- Security and correlation controls: Skylar AI strengthens credential handling. Trusted Certificates controls let administrators add their organization’s private certificate authority to establish trust for Model Context Protocol (MCP) servers whose certificates that authority issues. Entity Exclusion controls let owners exclude catch-all device groups and business services from advisory correlation to prevent unrelated advisories from merging. Remember – your data is secured to be sovereign to your organization and not shared with external foundation models or external parties!
“The first challenge in resolving an IT issue is often figuring out whether it’s a real problem, what’s causing it, and what actually needs attention,” said Michael Nappi, chief product officer at ScienceLogic. “With Skylar AI 2.8, we’re giving teams specialist agents that take on that initial investigative work, using live telemetry to deliver evidence-backed findings and recommended next steps. Operators can see the evidence behind those findings, and administrators control how and when agents run. The goal is to spend less time investigating and more time resolving the issues that matter.”
Skylar AI 2.8 follows three recent releases: Skylar AI 2.5 expanded secure deployment options in August, 2.6 added visibility and verification at the AI layer in September, and 2.7 introduced granular controls over agentic functions and MCP tools in October. The latest release extends those foundations into advisory investigation.
For more information about the ScienceLogic AI Platform, visit https://sciencelogic.com/.
About ScienceLogic
ScienceLogic delivers intelligence that accelerates outcomes through service-centric observability, AI-driven operations, and intelligent automation. Trusted by enterprises, service providers, and government organizations worldwide, our platform unifies visibility and intelligence with agentic AI so teams can move from understanding problems to resolving them with greater speed and precision. Founded in 2003, ScienceLogic serves customers globally and is recognized in The Forrester Wave™: AIOps Platforms, Q4 2026, and as a leader in the 2026 IDC MarketScape for Worldwide AIOps. Learn more at https://sciencelogic.com/.
View source version on businesswire.com: https://www.businesswire.com/news/home/20261110141037/en/
Media gallery
