we use a calculated field (using the addDateTime function) in the dataset that adjusts for the number of timezones between our local timezone and UTC (our datasource is redshift). Then we exclude the original timestamp from the dataset so that the users don’t by mistake use the original field in a new analyis.