Introducing our new "Analyze Alert Impact" agentic workflow for IT Operations Management Service Operations Workspace.
This workflow uses AI agents to provide IT Operators with critical information about alerts from 1st and 3rd party systems, helping operators identify the correct priority level, the blast radius, and the correct teams to be involved in the investigation.
How it works
When an alert comes into the Service Operations Workspace, it can be hard to figure out what the impact is or who needs to be involved. You may also need to notify a subject matter expert because observability tools are complicated, or you might not have the right access.
We’ve partnered with the great folks at New Relic and Kentik to bring online our first set of observability AI Agents. Through this integration, we are able to provide additional context and insight to speed up the investigation and triage process. Additionally, this insight is brought into the same workspace that operators are familiar with to reduce context switching.
New Relic AI Agent
For the New Relic analysis AI Agent, when an alert from New Relic comes in you can kick off this workflow via the Now Assist Panel by asking:
- Have there been recent deployments for this alert?
- How severe is this alert?
- What are the root cause theories for this alert?
- What else is impacted by this alert?
- What is the user impact for this alert?
- Which engineering teams should I pull in to investigate this alert?
- Who owns the impacted entity for this alert?
Our agentic workflow will autonomously connect to your New Relic instance and retrieve information about the alert. You’ll need to configure your agent with API credentials and the endpoint:
- For US customers: nrai-streaming.service.newrelic.com
- For EU customers: nrai-streaming.service.eu.newrelic.com
To send alerts from New Relic, please use the Integration Launchpad in the Service Operations Workspace. Additionally, to trigger this workflow, you can simply ask – via the Now Assist Panel – “What are the root cause theories for this alert?” or "What are the root cause theories for Alert12345?".
Kentik AI Agent
For the Kentik analysis AI Agent, you can kick off this workflow via the Now Asist Panel by asking:
- Are there network anomalies between the impacted services for this alert?
- Is there a DDOS attack in progress at the moment on the impacted services for this alert?
- Is there connectivity between the impacted services for this alert?
To configure your agent with API credentials and email address. To configure alerts from Kentik to flow into ITOM Event Management, please see the following KB.
What’s next
Reducing context switching, minimizing data silos, speeding up the investigation and triage process, and reducing outages / issues / incidents are all part of our roadmap. With the launch of these AI Agents and our Analyze Alert Impact workflow we are introducing new ways for IT teams to triage and troubleshoot issues in their environment.
As our environments become more complex – with issues crossing multiple monitoring and observability tools – it becomes more critical for teams to have the insights they need to resolve issues faster. With our agentic workflows, we are bringing those insights into ServiceNow to allow teams to interrogate 3rd party systems without losing the context of the alert or incident they’re investigating ultimately reducing the amount of time it takes to resolve issues.
We are in the early stages of our agentic journey so if there are any additional questions you’d like our agents to answer or vendors you’d like to see, we welcome all feedback from the ITOM community – feel free to comment below and tag me (@heidmo) in it.
