At the end of a sprint someone has to reconstruct what was done, what slipped and what changes for the people using the product. The agent works it out from tickets, commits and pull requests, and prepares a report and release notes that the team rereads before publishing.
A sprint's work is written in three different places: tickets, commits, pull requests. To tell it to someone who was not at the stand-ups, to a client or to management, someone has to put the pieces back together, and loses half a day every two weeks doing it.
Release notes are even worse, because they have to be written for the people who use the product, not for those who built it. So they come out late, too technical or not at all, and users discover new features by chance.
Closed and postponed tickets, commits, pull requests, descriptions and comments. Nobody has to fill in anything extra.
What was completed, what slipped and why, the risks that emerged, where things stand against the sprint goals. One page, not the list of tickets.
The same changes told from the user's point of view, grouped into new features, improvements and fixes. Internal details stay out.
They arrive as a draft in the team channel or the repository, get corrected and get published. The tone aligns with the examples the team approves.
These are the numbers we set in discovery, before writing any code. The starting value is measured on your data, and the prototype is judged against it.
The time spent preparing reports and release notes, before and after.
How many versions ship with their notes, and how long after the release.
How much the draft gets changed. If edits drop over time, the drafts are finding the right tone.
If tickets and commits are written in a hurry (“fix”, “misc”), the report cannot tell what nobody wrote down. That is usually the first thing to change: the team writes better tickets once it knows someone reads them.
Application logs and databases already show how people use them, but nobody has time to read them. AI reads them, writes a report and suggests what to improve.
An idea with AI at its core, a market to validate and no technical team. A real product in a few weeks, with real users, built to be able to change.
A free 30-minute call, with no commercial follow-up unless you want it.