Welcome Wednesday: Your Best Atlassian Intelligence and/or Rovo Tip!

[#1]

Happy Wednesday Everyone!

If you asked me the stance my company was taking on AI tools at the beginning of last year, the answer would be: “Looking, but not engaging.“ As of the start of this year, we have a team dedicated to reviewing AI tools (I’m not going to go into deal about that review process). However, those that are deemed low/lower risk are being “blessed“ for use.

This week Atlassian Intelligence and Rovo have both been given the greenlight to be enabled across our entire Atlassian stack. While communications are being prepared to inform users that this is being turned on, I fear that people may start drinking from a firehose on new features & capabilities they will have access to across Jira, JSM & Confluence.

So, my question for you this week: What is the best tip you have for me to pass along to my users about Atlassian Intelligence and/or Rovo? If you haven’t started using either yet, I’d love to hear what is holding your company back?

Have a great rest of the week!

[Barbara Szczesniak]

Happy Wednesday, @Jimmy Seddon!

My company has Jira Premium and Rovo is turned on there, so there may be some team members who are using Rovo in Jira.

We have Confluence Standard, so we can't turn on Atlassian Intelligence, and I don't see any Rovo buttons or anything, so I presume it's not turned on there.

I use Confluence to document our software. I do this by using the application, monitoring tickets for changes to features, and talking with product and ops team members about intended use. I don't know how AI could help me to do this, and I don't have time to investigate the possibilities. 

[Susan Waldrip]

Hi @Jimmy Seddon, similar to @Barbara Szczesniak 's situation, we use JSM Standard so won't have Atlassian Intelligence (think they'll trademark "AI"?!). While it's free to use now, I'm using Ask Rovo (Rovo Chat) to answer questions about how to configure things in JSM, especially when I can think of more than one way so Rovo provides options and pro's and con's for all. It's been helpful so far.

[Darin - Opus Guard]

Clean 👏 Your 👏 Context 👏 Window. 😄

Bad results = your team won't trust it, and ROVO is DOA.

Rovo and every other AI feature in Atlassian are only as smart as the content they’re reading… which means if your Confluence and Jira are full of old, wrong, or “we’ll-fix-this-later” pages, the AI will happily hallucinate based on all of it.

Think of it like giving Rovo a neat desk vs. a junk drawer: one gives superpowers, the other gives chaos. We’ve seen exponential improvements just by removing outdated, duplicate, or inaccurate content, and letting AI learn from what’s actually true today.

If you’re turning on Atlassian Intelligence across the org, this is the cheapest and fastest win you can give your teams.

For anyone wanting an easy way to automate that cleanup (policies, classification, retention rules, the whole thing), here’s what we use: Content Retention Manager for Confluence

Make the AI smart by giving it less to be confused about. 😉

[Dikla Tavor-Haimpur]

We're having a bit of a struggle with Rovo, to be honest.

Rovo chat is nice but as declared below, will happily mumble utter nonsense if your Confluence is not clean.

The real challenge is creating agents - so far, no one in our org has succesfully created a usable, popular agent using the Rovo studio and this makes people feel disillusioned. The main feedback I'm getting is that it looks very easy until you try using it in real life and see that it has challenges with customer fields, with teams that work differently, with the actual needs of a big, living organization.

Working to resolve this, but its not easy.

[Barbara Szczesniak]

Aah, @Darin - Opus Guard. Apparently, the bleeding edge technology of AI is not immune to the age-old problem of garbage in/garbage out. Though from things I've heard, you can get the garbage out part without the garbage in.

[Darin - Opus Guard]

@Barbara Szczesniak - Yea, it took me a moment to think about, but I think it's incorrect to say "garbage in" with this context. At some point, that PM, service agent, marketer, engineer, wrote that Confluence page with good intent and it was as well informed as it could be at the time. So I don't think the "in" was garbage at the time, instead, some content just ages like milk. Lessons are learned, regulations passed, customers fall out, technology changes, any number of factors and the all this user generated content goes from value to obsolete and eventually flat risky. 

Here's a slide lifted from our booth deck from Atlassian Team '25 that really captures it (I think).

But yea, to your point, we probably can't prevent new content from being created, nor should we. If anything, thanks to Rovo and AI in general, everyone should have a shot at creating much better content that can stand the test of time. It's still in everyone's best interest to remove the aging content, and many governance frameworks even ask us to do so.

[Artem Taranenko]

@Darin - Opus Guard - Appreciate you sharing this slide. It's prefect for my upcoming presentation to talk about importance of knowledge base governance, specifically with intention of using it as a data source for AI.