Is this similar to BigQuery's ANY_VALUE aggregate function? I found it to be extremely useful.