Use of Rovo in generating utilisation reports

[Febin Philipose]

Hi, 

I love using ROVO, and as a delivery manager, one of my key responsibilities is to find out if the resources in my team are under- or over-utilised against their planned capacity.

I have been prompting Rovo with the logic and calculation to compare the capacity (including deduction of planned leaves) in a three-week sprint in hours against the time spent or logged in a Jira ticket, with an expectation from the team to ensure they have atleast 80% time logged in their assigned tickets within a sprint.

I was able to extract the report in confluence but the biggest challenge is:

• Rovo takes a lot of time to identify and compare the capacity v/s time spent, and at times pauses, and if you prompt it again, re-validates the calculation again, taking almost 15-30 minutes again to re-process again. So I need to be on the Confluence page till this is analysed.

Do you think having an agent to do this would be helpful? The challenge here is, every time I need to extract the time spent in csv and attached it to Rovo. I wish I could create a ROVO custom agent with the time-track gadget in-built to avoid this manual csv upload process and ensuring utilisation report updates automatically weekly, monthly etc. That could be really helpful.

Reaching out to the community here to help if is there is a solution.

Kind regards,

Febin

[Tomislav Tobijas]

Hey @Febin Philipose ,

Having an agent for this would definitelly be helpful. We do have an effort estimation agent that looks at previous estimation (which are all Confluence pages) and based on that and the user requirement, the agent provides a detailed breakdown and rough estimation per milestones. 👀
Now, the prerequisite for that is to have all that data in Confluence as separate Confluence pages, which all use a defined label.

Anyway, in your particular case, I would suggest building this agent and then try using it in Jira (or Confluence) automation. 

🤔 What you could try is breaking down this agent you built into an agent with subagents.
You can create specific scenarios for "Capacity Calculation" and "Time Spent Analysis" so the agent doesn't have to re-validate the entire logic for every prompt.

Note that I haven't played that much with subagents so that would need to be tested a bit.

Cheers,
Tobi

[Febin Philipose]

Hey Tobi,

Thank you — this is really helpful.

I honestly hadn’t considered breaking this down into smaller agents/subagents, but the logic makes sense. Thinking about it now, one of the biggest challenges is probably Rovo trying to re-process both capacity and time spent every single time instead of treating them separately.

The automation piece is what really caught my attention as my biggest pain point today is manually exporting Jira time logs to CSV and re-uploading them for analysis.

What I’m really trying to get to is a way of automatically comparing sprint capacity (including planned leave) vs actual logged effort and surfacing utilisation insights without manual effort.

I’ll definitely experiment with this approach.

• Curious — have you seen anyone solve this well with changing Jira worklogs, or is this still something people are figuring out with Rovo?

Really appreciate the guidance 👀

[Tomislav Tobijas]

Curious — have you seen anyone solve this well with changing Jira worklogs, or is this still something people are figuring out with Rovo?

I'd probably say everyone is still experimenting. 🧑‍🔬 (myself included)

I don't recall anyone having this specific use case, but someone might chime in here with their experience. 

I'm planning to do some specific tests related to time spent analysis in the next or 4th quarter of this yet 👀
My idea was to create specific reports (dashboards or Confluence pages) once a project ends, and then, based on that, feed my effort estimation agent for future estimations. 
Again, this is all theoretical, but I might write an article or discussion about it once it test it out and see how it operates. 

[Febin Philipose]

That’s really interesting, Tobi — especially the idea of using historical time spent data to improve future estimations.

From a delivery perspective, I can definitely see value there. One thing I often notice is that estimations gradually become less reliable when team composition, complexity, or dependencies change, so having historical delivery patterns as context could make planning much more grounded.

It also makes me wonder whether Rovo could eventually move beyond analysing historical effort and start surfacing potential estimation or utilisation risks earlier in the delivery cycle.

I’d be genuinely interested to see how your experiments turn out — especially around time spent analysis. Please do share when you test it, as I’d love to compare notes while continuing to experiment with utilisation analysis on my side 👀