The tabbed part is easy to explain but can be complex to execute.
Method 1 (sloppy)
Render a set of layers for each tab. Make them visible and invisible when user clicks "tab" button.
Make so you have a function to draw, and a function destroy each "tab" buttons elements.
User clicks on a tab, it destroys all layers, then draws the chosen layer.
Method 2 is harder because you have to map where everything is before you destroy the tabs elements.
Check out the Families function in Construct 2... You could have a family for each tab.
You can also use containers to help out.