A practical guide to Atlassian's Rovo Agent Profiles

So, Atlassian has officially jumped into the AI game with Rovo, a new platform that threads artificial intelligence through tools you probably use every day, like Jira and Confluence. The big stars of the show are Rovo Agents, which are basically customizable AI teammates built to handle tasks and, hopefully, make life easier. They sound pretty impressive, and in many ways, they are.

But, like with any shiny new tech, there’s a bit of a learning curve. What are these agents, really? How do you keep them from running wild? And maybe most importantly, what are the gotchas, the hidden costs and limits, you should know about before you go all-in?

This guide is here to cut through the hype. We'll walk through what Atlassian's Rovo Agents and Rovo Agent Profiles are all about, how to get them set up, and the big questions around pricing and control that every team leader needs to ask.

What are Rovo agents and Rovo agent profiles?

Alright, let's break this down. In simple terms, Rovo Agents are AI helpers you can build or use right inside your Atlassian tools. They’re meant to handle specific jobs, like summarizing a monster Confluence page or tidying up your Jira backlog. Atlassian gives you more than 20 pre-built agents to start with, but you can also make your own from scratch.

So, what about Rovo Agent Profiles? Think of a profile as the agent’s resume or ID badge. It’s a single place where you and your team can get the scoop on any AI agent.

An agent profile tells you:

• Who made it: So you know who to bug with questions.

• What it's supposed to do: This covers the specific prompts and rules guiding its behavior.

• Where it gets its info: You can see if it's pulling from certain Confluence spaces, Jira projects, or even a connected Google Drive.

• What it's allowed to do: This lists its "skills," like whether it can create a Jira ticket or edit a document.

Getting a handle on these profiles is pretty much ground zero for using and managing Rovo agents without causing chaos.

Key features and capabilities

Rovo agents are pretty flexible, but their real magic comes from how you set them up and what you connect them to. It all boils down to a few key ideas.

Understanding pre-built vs. custom agents

Atlassian gives you a bunch of pre-built agents for everyday tasks, like the Issue Organizer or Release Notes Generator. They’re a solid way to dip your toes in the water. But the real fun begins when you start building your own. You can whip up "No Code" agents by just chatting with a bot to give it instructions. Or, if you have developers, they can build more advanced "Custom Agents" on Atlassian's Forge platform.

Fueling your agents with the teamwork graph

The brain behind Rovo is something Atlassian calls the "Teamwork Graph." It's a fancy term for a system that understands how all your people, projects, and documents are connected. By plugging in sources like Confluence or even outside apps like Google Drive, you're feeding your agents the context they need to give smart answers and get things done.

The catch: Jack-of-all-trades vs. a master of one

Here’s something to think about: Rovo is built to be a generalist. It’s a company-wide AI that can do a little bit of everything, summarize meetings, draft docs, organize backlogs. That's great for versatility, but it might not have the deep expertise needed for specialized jobs like customer support or IT service management (ITSM). When the stakes are high, you need an AI that gets the tiny details of a support ticket or an IT incident.

This is where you'll see purpose-built tools shine. An AI platform like eesel AI, for example, is built from the ground up for support and ITSM. It doesn’t just link to your help desk; it actually learns from your past tickets to pick up on your team's tone, understand common problems, and know what a good resolution looks like. Rovo gives you a box of parts, while specialized tools hand you a solution that’s already fine-tuned for the job.

A workflow diagram illustrating how a specialized tool like eesel AI automates the customer support process, a key feature of Rovo Agent Profiles.

Creating and managing Rovo Agent Profiles: What to expect

On paper, making a custom Rovo agent sounds easy enough. You chat with a bot, tell it what to do, point it to some documents, and set its permissions. But as anyone who’s tinkered with new tech knows, theory and reality can be two different things. Here are a couple of hurdles you should probably expect.

Get ready to tweak, and tweak again

Early users have reported that getting a Rovo agent to behave reliably takes a lot of trial and error. You have to be super specific with your instructions, and even then, the agent can get things wrong. You might tell it to summarize issues by epic, and it just… makes up epics that don't exist.

Reddit

Why your documentation needs to be spotless

An AI agent is only as smart as the information it can access. If your Confluence pages are a mess of outdated info, your Rovo agent is going to give you messy, outdated answers. A lot of teams find they have to kick off a huge documentation cleanup project before they can even get started, which can really slow things down.

A less painful way to get started with AI

These hurdles point to a big missing piece in Rovo's setup: a safe way to test things out. This is where a platform like eesel AI does things differently.

• Test without the stress: Before you unleash anything on your customers or team, eesel AI has a simulation mode that lets you test your AI on thousands of your past support tickets. You can see exactly how it would have replied, get solid predictions on how many issues it can solve, and tweak its behavior in a sandbox where nothing can break.

A screenshot of the eesel AI simulation mode, which helps in managing Rovo Agent Profiles by allowing risk-free testing.

• Launch in minutes, not months: Forget about waiting for a gradual rollout or spending half a year cleaning up docs. eesel AI is self-serve, so you can connect your help desk, click a button, and have a working AI assistant up and running in minutes.

• It learns from your actual work: Don’t have a pristine knowledge base? No worries. eesel AI is built to learn directly from your team's past conversations. It automatically picks up your solutions, tone, and workflows, making your messy, real-world data its greatest asset.

This approach lets teams get up and running quickly, seeing real results without the massive upfront effort that some general tools require.

Atlassian Rovo pricing explained

Let's talk about the big question on everyone's mind: what is this all going to cost? For a lot of teams, the biggest hesitation with Rovo isn't about what it can do, but what it'll do to their budget. Atlassian's pricing plan, or lack thereof, has left a lot of people scratching their heads.

Right now, Rovo is being offered to Premium and Enterprise cloud customers as part of their existing plans. But here's the catch, Atlassian has made it clear this is just a "promotional period." Sometime in early 2026, they'll switch to usage-based pricing. This has teams worried that they'll get hooked on a tool only to have the financial hammer drop later, making it way too expensive to keep.

To make things even more confusing, the current usage limits are incredibly tight. Based on Atlassian's own documentation, an Enterprise user gets about seven "engagements" a month. If you're a busy support agent or a project manager, that’s... nothing. You could burn through that in a single morning.

It's tough to build your workflows around a tool when you have no clue what it will cost you next year. How can you possibly budget for that?

Finding a more predictable path

This is another reason why a specialized platform can be a breath of fresh air. Most modern AI tools were built from day one with clear, upfront pricing.

Take eesel AI, for instance. It offers simple pricing based on the features you need and your interaction volume, with no surprise fees for every problem it solves. You know exactly what your bill will be each month, so you can grow without dreading a massive invoice.

A screenshot of eesel AI's pricing page, showing a clear and predictable pricing model, which is an important consideration for Rovo Agent Profiles.

Here's a quick look at how the two pricing approaches stack up:

If you're a team that needs to plan your budget, a platform with transparent pricing just makes more sense. It lets you focus on the value you're getting, not on guessing what the cost might be.

Are Rovo Agent Profiles right for you?

There's no doubt that Atlassian Rovo agents are an interesting new piece of the puzzle, giving you a way to build AI helpers right into your existing workflows. And the idea of Rovo Agent Profiles brings some much-needed order to managing them.

But it's also clear that the platform is still in its early days. If you're thinking about jumping in, you'll need to balance its flexibility with some real-world challenges: a steep learning curve, the need for perfect documentation, and a pricing model that's a giant question mark.

For teams working in customer support, ITSM, or running internal help desks, a more focused, proven solution might be a better bet right now.

If you're looking for an AI platform that's actually built for support, gets going in minutes, and has pricing you can understand, why not give eesel AI a try? Get started with eesel AI for free and see what it can do for your team.