How best we can use Rovo & AI in Enterprise Service Management

[Chethan GR]

I was wondering how organizations are leveraging Rovo and AI in Enterprise Service Management. Has anyone implemented it or explored use cases already? I’d be interested to learn from your experience and understand the best practices.

[Christopher Yen]

Here are some use cases we've recently implemented specifically within JSM

• Change Risk Assessments
  • Agent that provides an initial risk assessment when change requests are submitted in our JSM project to provide IT, Security, Compliance analysis for a change, feedback for the submitter on any aspects where the change plan can be improved and also using our checklist where risk points are associated with each option providing a score and risk level. This is triggered by automation and within the rule we use the agent response to comment on the work item and write to some fields. 
• Virtual service agent
  • We're still figuring out how incorporate our forms and conditional + custom field experience with Rovo in IT so still in sandbox on this one but we've seen some success with our non IT teams that want to intake work items via Slack + virtual service agent as long as their request types are straight forward and simple. Rovo can definitely work for us in IT but it's just going to take more trial and error with its configuration to ensure a good user experience. 
• Some tier 1 support after ticket creation.
  • We have some scenarios where tickets get created and we can fully automate either the license assignment or redirection but it depends and we're using Rovo to determine whether the request is straight forward or is a case where human intervention is needed and if so to transition the ticket to escalated, but if not we write a specific response to a custom field that triggers another automation to grant a license to the customer (workato recipe > okta group) or provide a canned response 

In terms of best practices I'm not really sure, Rovo is constantly evolving and there's always ways to improve the instruction or need to adapt as Rovo comes out with new skills but Atlassian has their own doc on best practices for Rovo here https://support.atlassian.com/organization-administration/docs/best-practices-for-atlassian-intelligence/

Specifically for ESM I would say having a knowledge agent that has access across your Confluence site is one of the better and easy to set up use cases. 

[Chethan GR]

Thanks @Christopher Yen  for sharing your insights, it's really helpful!

[Arkadiusz Wroblewski]

Hello @Chethan GR 

@Christopher Yen  examples are a solid start. From an ESM perspective, it is usually better to avoid the "AI everywhere" approach at the beginning. Instead, focus on a single, high-volume service area where your documentation and processes are already well-defined. Rovo is essentially a force multiplier for your existing knowledge, so if your forms and articles are currently disorganized, the AI will likely just make those weaknesses more visible to your users

The ultimate goal should be reducing repetitive manual work and helping users find answers faster, rather than trying to replace your agents entirely. Success in ESM usually comes down to whether the AI actually helps the team move through the queue more efficiently without losing that human touch on complex issues.