Introduction

Key Terminology

• “🚀 Action” denotes a setup step for the presenter.
• “📣 Narration” denotes what the presenter will say.
• “ℹ️ Note” denotes where the presenter may need to deviate from this demo script or add supplemental comments.

• Application: IBM CloudPak for AIOps brings together the capability to group resources from different data types into applications. Clients can flexibly define an application to meet their business needs. With applications, you can obtain an integrated view of resources to understand inter-dependencies.
• Event: A point-in-time statement in CloudPak for AIOps that tells us that something happened somewhere in a client’s environment. It tells us what happened, where it happened, and when it happened. An event does not have to be exceptional or actionable, it can simply tell us something has happened.
• Alert: An alert in CloudPak for AIOps represents an abnormal condition somewhere in an environment that requires resolution. It tells us what is happening, where it is happening, and when it started to happen. It may be informed by one or more events. It has a start time and end time.
• Story: A story in CloudPak for AIOps represents an outage or reduction in service which is currently impacting customers and requires rapid remediation. It is created based on one or more trigger alerts that indicate the outage or reduction in service. Any alert of severity Major or Critical will act as a trigger alert. Other alerts that share the same cause may add context to the story.
• Incident: An incident in ServiceNow is an event of interruption disruption or degradation in normal service operation. An open incident in ServiceNow implies that the customer is impacted, or it represents the business risk.
• Topology: A topology is a representation of how constituent parts are interrelated. In CloudPak for AIOps, an algorithm analyzes how the event nodes are proximate to each other and groups them into a topology-based correlation.

Use Case

1. Pulls data from various IT platforms: IBM CloudPak for AIOps monitors incoming data feeds including logs, metrics, alerts, topologies, and tickets, highlighting potential problems across incoming data, based on trained machine learning models.
2. Utilizes AI and natural language processing: An insight layer connects the dots between structured and unstructured data, using AI and natural language processing technologies. This allows you to quickly understand the nature of the incident.
3. Provides trust and transparency: Using accurate and trustworthy recommendations, you can move forward with the diagnosis of IT system problems and the identification and prioritization of the best resolution path.
4. Resolves rapidly: Time and money are saved from out-of-the-box productivity that enables automation and utilizes pre-trained models. A “similar issue feature” from past incidents allows you to get services back online for customers and end-users.

Demonstration flow

1. Scenario introduction
2. Trigger problem situation [In the background]
3. Verify the status of the Robot Shop application.
4. Understanding and resolving the incident
   1. Login to AI Manager
   2. Open the Story
   3. Examining the Story
   4. Acknowledge the Story
   5. Similar Incidents
   6. Examine the Alerts
   7. Understand the Incident
   8. Examining the Topology
   9. [Optional] Topology in-depth
   10. Fixing the problem with runbook automation
   11. Resolve the Incident
5. Summary