You are right. There is no pseudo column "rowid" in teradata.
If you want to eliminate duplicate records (basically all columns having same value in more than one record), then you can create a SET table. Do simple insert into select * from. It will silently eliminate duplicates. Note that SET table can not allow duplicate records, MULTISET table allows duplicate records.
But in case you specified, you want to get latest record based on set of columns if they repeat. You can do this using ROW_NUMBER() function. I am assuming you have a modified_ts in your table to identify latest record. If so, below query can be used to get latest records.
select customerid, customename, customeraddress, ROW_NUMBER() OVER(PARTITION BY customerid, customername ORDER BY modified_ts DESC) rn from customer QUALIFY rn=1