Data workspace went GA a few days ago. While there is no doubt that this is the future for creating Dataverse data models, it does have two irritating shortcomings. The first issue is that you cannot save and edit your model. Once you click “save and exit,” the tables are created and your model is discarded.
The second issue is that you cannot add existing tables to your workspace. If you try to and trick it Copilot by asking it to create a relationship between your table and the “contact” table, it will instead create a new table named “contact.”
Those glitches aside there is no doubt that this is the way forward for Dataverse data modelling. The user interface is simple to navigate, and your ever-present companion, Copilot, is available to assist you in creating your data model.
The view data option allows you to visualize how your data will look and easily modify your columns properties. Much quicker than using the Power Apps Maker Portal especially if you are new to Dataverse.
Data Workspace also makes it easy to create tables from SharePoint Lists or Excel. I have used it to move data across Dataverse environments by exporting from the source to Excel and importing into Data Workspace in the target. You will get some data validation errors for the hidden columns but you can simply delete those columns.
Data Workspace is solution aware and if you use it from within a solution it will create your table with the solution’s publisher prefix. If you use it outside a solution if will use the prefix from your default publisher so it is always good practice to assign a preferred solution in an environment.
If you open the Data Workspace from Apps you get the option to save your changes and create either a Canvas or Model app.
I am looking forward to the next release of Data Workshop which hopefully will fix the niggles I mentioned and bring full ERD capabilities to Dataverse. If you want to create a Dataverse ERD diagram currently you need to use external tools some examples here.