What is Sprint Insights

Sprint Insight is a way of checking whether a team is working mostly on features or is doing maintenance work. Sprint Insight uses a number of orchestrated machine learning algorithms that perform analysis of project management software, extract issue tickets from it and predict whether a given ticket describes a feature or a not-feature task.

Sprint Insights connects to your ticket monitoring system (Jira) and receives updates about all user activities. After an issue ticket is created, the initial analysis is performed and the type of the ticket is determined. Next time when the ticket is modified a new analysis is done and the classification is adjusted. This process ensures that the Sprint Insight system stays up to date with all updates that happen and every time uses all the available data to make the most accurate predictions. Information about all made predictions is stored and then monitored and adjusted based on the user's input.

Setting up Sprint Insights

To set up Sprit Insights Agile Analytics needs to be connected to your ticket feed.

Connect Jira: https://zensoftwarenl.atlassian.net/wiki/spaces/AGILEX/pages/2294841385

Connect Azure DevOps: (coming soon)

After connecting your ticket feed Agile Analytics starts gathering and displaying data on the Sprin Insight Page.

How to train the AI model

The machine learning model used for issue ticket classification can be trained to further improve its predictive capabilities. In order to correct an erroneous label do the following:

Select a ticket that has a wrong label:

Select the incorrectly classified ticket
Click on the Select Action dropdown menu and select Train.

the Select action dropdown menu
In the popup select the correct label.