Rovo Agents in Real Life

[#1]

AI assistants often promise a lot, but in real-world work, they are often outside the tools where the real work occurs. Rovo Agents are different.

Rovo Agents are integrated into Atlassian (Jira and Confluence) tools, rather than existing as a separate chatbot in another window. They are designed to understand your work, your projects, and your context, and to help you move faster by providing real answers to real questions.

Rovo Agents require no special commands or technical skills. You only need to ask in simple terms, and the agent does the hard work of searching, contextual interpretation, and sending helpful findings.

In this article, we discuss what Rovo Agents are and how teams can build and use them in practice.

What Are Rovo Agents?

Rovo Agents are AI assistants designed to work with Atlassian products and are task-oriented. They assist in a particular type of work - searching, summarizing, reporting, or analyzing information across Jira and Confluence.

Rovo Agents handle natural-language queries. They understand context (projects, timeframes, users, permissions) and can provide practical outputs, rather than raw data. They are right where teams are already working.

Rovo consists of three related capabilities:

• Rovo Search - search results in Jira, Confluence, and related tools.
• Rovo Chat - enables you to pose questions in a conversational manner.

• Rovo Agents - specialized assistants that are supposed to solve a specific problem.

You can consider Rovo Agents an AI colleague who has a specific role.

How to Create a Rovo Agent in Jira (and What Makes a Good One)

Building a Rovo Agent is not about AI magic. It is about having one clear, repeatable problem to solve that real people deal with on a daily basis.

The best Rovo Agents are less bot-like, and more like teammates who understand precisely what they are expected to do.

Follow these steps to create a Rovo Agent in Jira: 

1. Open Rovo in Jira

• Open Jira and at the top of the page, click Ask Rovo.
• Open Menu (☰) and choose View all agents.
• Then, click Create agent.

2. Give the name to your agent

• Use a role-based and purpose-driven name. The agent’s name should tell users exactly when to use it.

3. Write the instructions to the agent

• Specify the role of the agent (e.g., “Project Status Assistant” or “Marketing Campaign Assistant”).
• Define the target user (e.g., PMs, marketers, engineers).
• Focus on one core problem.
• Specify the Jira info to use.
• Tell the agent to automatically understand context like projects, time ranges, and user permissions.
• Demand brief summary, then details.
• Prefer what matters most (blockers, high priority, risks).
• Specify what the agent should avoid (e.g., irrelevant data, unnecessary detail, assumptions).

4. Add conversation starters

• Use conversation starters to show the users how to ask questions and what the agent is capable of doing.

5. Define the context

• Rovo is permission aware and recognizes projects, dates, and typical Jira fields automatically.
• Make the agent concentrate on the correct projects and leave Rovo to determine the context, unless you require strict rules.

6. Test the agent

• After creating the agent, ask one of the starter questions.
• Refine instructions if responses are too long or too technical.

Example: Creating a Rovo Agent to Analyze Monthly Support Tickets

This example shows how it is possible to build a Rovo Agent that processes support tickets over the month and delivers insights that can help project managers and marketing teams.

The main instructions for the agent: 

• Goal: The help teams see the number of created support tickets in a month, the most frequent issues, and the areas where customers had the most problems.
• Target users: PM, marketing managers.
• Project scope: Jira project - SUPPORT.
• Agent name: Support Tickets Analyst.
• Instructions:

You analyze support tickets for one Jira project over a selected month.

Focus on:

- Tickets created in that month

- Common themes and request types

- Recurring issues and volume spikes

Use clear, non-technical language.

Start with a short summary, then explain why the insights matter for project managers and marketing teams.

• Conversation starters:

“What were the main trends in the SUPPORT project last month?”

“Which issues were most frequent in the SUPPORT project last month?”

“Were there ticket volume spikes in the SUPPORT project last month?”

Here is an example of how such an agent works: 

Why this agent is useful:

• For PMs: allows spotting recurring issues, understanding support risk, and prioritizing fixes.
• For marketing department: allows identifying customer pain points and use them in messaging and positioning.

Changes Report Assistant as an Example of an App-Powered Rovo Agent

An effective method of learning about Rovo Agents is to consider an actual example created by a Marketplace app.

Changes Report Assistant is a Rovo Agent offered by Issue History for Jira app (SaaSJet). It makes teams to know what has changed in Jira projects using easy, natural-language questions.

When a query is made by the user, such as: “Show me all the changes made to the MARC project last week”, the agent automatically understands your intent and generates the link to the needed report. 

The report will look like this:

It can be easily exported in Excel or CSV. 

The Changes Report Assistant eases the work of project managers by enabling them to:

• Check out the latest changes in projects.
• Spot priority or scope changes.
• Prepare status updates and audits more quickly.

Try Issue History for Jira to explore the Changes Report Assistant and see Jira work item changes clearly 🚀

Summing Up

Rovo Agents assist teams in easier usage of Jira since it allows asking questions in natural language and receiving straightforward answers. They are most useful when they focus on one specific task and support a clear role.