recursive function iterate over python object

Question

I have developed a system with different classes. Some classes are attributes of other classes, such as:

@dataclass
class MyFirst:
   name: str
   numb: float
   time: pd.Series
   my_second: MySecond = field(default_factory=lambda: MySecond()

@dataclass
class MySecond:
   name: str
   date: pd.Series

I have this structure for about 2-5 sub-classes being the attribute of another class. Now I would like to write my object:

system = my_first()

To a dictionary. I create a tuple with all my classes at first

my_classes = (MyFirst, MySecond,...)

Then, I iterated over the highest-class

dict_system = {}
for attr in dir(system):
        if not attr.startswith("__") and not callable(getattr(system, attr)):
            if isinstance(getattr(system, attr), (pd.Series, pd.DataFrame, float, int, str)):
                dict_system[attr] = getattr(system, attr)
            elif isinstance(getattr(system, attr), my_classes):

In the elif, I need him to jump again and iterate the new subclass, which would again be a dictionary, creating a nested dictionary. Finally, it would look like this:

dict_system = {
"name": "some_name",
"numb": 3,
"time": pd.Series([0,1,2,3]),
"my_second": {
         "name": "second_name",
         "date": pd.Series([date_1, date_2,...])
     }
}

Is there a way to build this in a recursive function so I can have as many subclasses as I like, and they would appear in a nested dictionary?