Hi everyone! đź‘‹
In 2026 when an alert fires, responders still have to piece together the context behind every alert by hand. Working out what it means—and what to do next—still takes too much manual effort. Responders switch between alert details, monitoring tools, runbooks, and chat threads to piece together the puzzle. Before they can act, they need to answer basic questions: What’s happening? Has this happened before? Where should I investigate first? That work takes time while a potential service issue is unfolding.
Alert Investigation brings AI-native investigation directly into your alert response workflow in Jira Service Management and Slack (coming soon). Start an investigation from an alert or alert group in Jira Service Management, or from an alert notification in Slack. Rovo provides initial findings to help you understand the alert and focus your next steps. Ask follow-up questions to explore further, and connect your observability tools through Model Context Protocol (MCP) to give investigations richer context and get better results.
The goal is to reduce the time spent gathering context and help responders get to a useful next step sooner, wherever they’re already working.
We’re rolling out Alert Investigation as an Open Early Access Program (EAP) feature over the next 4–6 weeks.
See an early demo of alert investigation in action:
https://www.loom.com/share/38b7202888294bb4ac1636c6d046222b
Who is this for?
Alert Investigation is designed for:
- On-call engineers and SREs who investigate alerts and want help getting to a useful starting point.
- IT operations teams that respond to alerts in Jira Service Management and collaborate in Slack.
- Teams using observability tools that want to explore how connected context can improve investigations.
Availability: This feature is available to all Premium and Enterprise licensed customers.
Get better results by connecting an MCP
Alert investigations become more useful when they can draw on relevant context from your tools. Connecting an observability tool through MCP gives Alert Investigation additional context to work with and improves the results.
MCP, short for Model Context Protocol, provides a way for AI capabilities to connect to external tools and data sources.
MCP connections are supported for the following providers: New Relic, Dynatrace, Splunk, Datadog, Lansweeper, Sentry, Honeycomb, and Coralogix.
What to expect during early access
We’ll enable the Open EAP progressively over the next 4–6 weeks, so availability may differ across sites during the rollout.
This is an opportunity to try Alert Investigation in your workflow and help shape its development. Review the findings alongside your team’s operational context and use follow-up questions to explore anything that needs clarification.
We’re especially interested in learning:
- Did the investigation help you decide what to do next?
- Which findings were useful, and what context was missing?
- How did connecting an MCP affect the results?
- Was it easy to enable the feature and start investigating?
Share your thoughts and feedback
You can share your questions and feedback in the comments or get in touch with our support team. We look forward to hearing how Alert Investigation works for your team!