When uploading a new CSV file, the system will not refresh the materialized views correctly if there is corrupted data. Sometimes this feedback is only available when refreshing the materialized views, namely when executing the SQL queries. Therefore, it would be nice to execute them isolated, before populating the achilles_results table with damage information.