Are you familiar with Event Sourcing? Have you used it?

Event Sourcing is an architectural pattern where all changes to the system's state are stored as a sequence of events, rather than just the current state. Instead of storing only the final data, the system keeps a history of changes, allowing for state recovery at any point in time and analysis of the history.

In Python, Event Sourcing can be implemented, for example, using libraries like eventsourcing or by creating your own layer of events and aggregates.

Here's an example of a simple event and aggregate:

class Event:
    def __init__(self, data):
        self.data = data

class AccountCreated(Event):
    pass

class Account:
    def __init__(self):
        self.events = []
        self.balance = 0

    def apply(self, event):
        if isinstance(event, AccountCreated):
            self.balance = event.data.get('initial_balance', 0)

    def create(self, initial_balance):
        event = AccountCreated({'initial_balance': initial_balance})
        self.apply(event)
        self.events.append(event)

I have used Event Sourcing in projects to ensure full traceability of changes and to simplify state rollback.