How AI Can Help Teams Estimate Smarter in Jira — Thoughts from a CEO Working with Agile Tools

[#1]

Hi everyone.
As someone leading a product team focused on Agile estimation, I’ve been thinking a lot about how AI can actually support — not replace — the human part of sprint planning.

In most teams I’ve worked with, the toughest part isn’t assigning numbers — it’s aligning perspectives, removing bias, and keeping estimations consistent across sprints.
We started experimenting with ways AI could learn from past Jira data (like historical story sizes, completion times, and sprint velocity) to offer contextual suggestions.

The idea isn’t to automate decisions, but to give teams better starting points for discussion.
It’s been interesting to see how this approach helps reduce noise and keeps planning sessions shorter and more data-driven.

I’d love to hear your thoughts — do you see AI having a real role in sprint estimation, or do you prefer to keep that process fully human?

(For context — I’m the CEO of GoAgile.ai
, where we experiment with AI-assisted Agile tools, but I’m asking here mainly to learn from other practitioners’ experiences.)

[Tarang]

May be the real value of having an AI system here would be to surface Risks and any blind spots with the problem scope stated for story.  As for past estimation data may be only useful to give a sense of order of magnitude given that a team isn't solving the same problem over and over, for that matter the team has gained new experiences and knowhow since then and so their estimates are likely to be informed by this vs. calcifying to a past estimate.

Giving an estimate, and a number is easy, the real problem is surfacing risks and unknowns that are best done through conversation exchange between members of a x-functional with diverse experiences.