A way to prevent this in the future would be to use a “find or create” microflow, something like this:
For your current issue, if the duplicate records are identical, it should be possible to write a fix flow that removes any duplicates, for instance:
Here i need to remove the duplicate records..how could i do that?