How would you use a Python dictionary keyed by customer_id to keep only the latest record per customer, overwriting a record only when the new updated_at timestamp is higher?

You can iterate over the list of records once, maintaining a dictionary that maps each customer_id to the record with the most recent updated_at. For each incoming record, check if its customer_id is already in the dictionary. If it isn’t, simply add the record. If it is, compare the updated_at of the existing record with the new one; if the new updated_at is greater, replace the entry. This guarantees that after the pass, the dictionary contains only the latest record for each customer. The algorithm runs in O(n) time, where n is the number of records, and uses O(k) space, where k is the number of unique customers. Example code:

latest = {}

for rec in records:

cid = rec['customer_id']
if cid not in latest or rec['updated_at'] > latest[cid]['updated_at']:
    latest[cid] = rec

# latest now holds the newest record per customer.

This approach is efficient for both batch and streaming scenarios and can be extended to handle more complex merging logic if needed.