In my opinion it's only a first dive into what's going on with your target variable.
You should assess the results and build a more informed model as you would have done without running automated analysis and and most of all, having knowledge about the data and the context.
To make clear my last point: Imagine that you as a english speaker get the assignment to explain the house prices for Tokyo, but you have the data table completely in japanese without labels and translations.
They only tell you what is the target variable.
That's an use case for automated analysis.