Introduction to Tempo Machine Learning Models

Tempo is moving towards automating and making life easier for our end users. The first area of automation is the Activity Feed. You would have already noticed suggestions in My Work calendar and list view. This feature is now being enhanced using the Machine Learning model. You will see a mix of machine learning model suggestions and other issue keys that you previously used to log time in the timesheet. The predictions are based on the Event Title, Event Description, and historical logged time of all the Tempo timesheet cloud users.

With the introduction of the machine learning model, the following features are available:

Suggestions in Issue Picker :

The machine learning model predicts based on previously logged activities in your timesheet. As Tempo is embarking on the path of using Machine learning models instead of rule-based engines, to provide more precise predictions and ease the time entry process.

When you click the issue picker, there will be a maximum of 4 suggestions predicted by our learning models which are denoted by a lightning bolt icon. You will see a mix of machine learning model suggestions and other issue keys that you previously used to log time in the timesheet.

To find and select an issue from Log Time form:

Missing Issue Keys :

The events coming from Google or Microsoft calendar integrations with no Jira issue key specified in the title of the event are displayed as Incomplete activities in the My Work view.

With machine learning, you will now see fewer incomplete activities as the model predicts the Jira issue key for some of them. When you click the issue picker, there will be a maximum of 4 suggestions predicted by the machine learning model which are denoted by a lightning bolt icon.