From Raw data to visualization

In order to bring your cloud data into Glean, you need to give us a little information so that we can build a model to base our visualizations off of.

The biggest design challenge in this was making this process powerful enough for folks who are familiar with SQL workbenches, while giving folks who don't have in depth knowledge access to their organization's data.

We landed on this workbench that allows you to select a table to start visualizing data. Once chosen, Glean will bootstrap your data into a basic model. This gets you to a visualization immediately, but also gives you a starting point for how to understand the model building process.

We used the principle of progressive disclosure in order to focus the user on what they need to do in the moment. This is done by only expanding the 'model tray' once you've clicked the 'build model' button.

For folks who want to write SQL, we've created a workbench that will allow users the query the table directly. The process for adding a query to the table follows the same pattern for adding in a table without SQL. Clicking the 'Build Model' button will bootstrap the model with what Glean needs to create a visualization and the user can add what other columns they see fit.

So much more to say about this project and tons more to be done! I'm really excited to have it out there in the world and folks playing with it.