Build a subnested dictionary from a Pandas DataFrame

Question

Alright, this is perhaps something rather simple. I want to know how to generate a nested structure of objects(recursive dictionary). Something like this:

{
   top_parent_child:
      child: {
         child:{
            child:{}
         }
      }
}

There is a point where the object has no more child objects. Now, to be more specific, I'm getting this data from an excel file and processing it with Pandas. The file has the following shape:

object_id	parent_pbject
ABC123
BCD123	ABC123
BCD345	ABC123
BCD234	BCD123
BCD567	BCD234
ABC789	BCD234
ABCD12	ABC789

I want this:

{
   ABC123: {
      BCD345: {}
      BCD123:{
         BCD234:{
            ABC789:{
               ABCD12:{}
            }
            BCD567:{}
         }
      }
   }
}

One child might have multiple sub-childs and so on. There are many suggestions I have received for this, like exporting a JSON using Pandas and then run through the entire things to recursively build the dictionary. But IMO the best option is to use Pandas with "groupby". For brevity, I used df.groupby(['parent_object']) after doing all the required filtering in the file because it is pretty big. After all that is done, I export the result as a dict, but it doesn't quite have the structure I'm looking for.

Do you have any better approach for this with Pandas? I'll appreciate any insight.