We ingest data from a variety of source systems, place them in S3 buckets, and sync those data into a warehouse using FiveTran. In one instance we download reports from Alexa Developer Console and can't control the names of columns in the downloaded .csv file. In this example, we have columns named 'TurnOn' and 'turnOn' to indicate commands a user makes to their Alexa device. Each of these columns contain values we want to retain, but Fivetran won't sync both due to the way it handles the case of the text - it sees them as duplicates. There are more instances like these where we want to ingest unadulterated data from sources that we don't have the power to change. Any enhancement that retains the case of the column header or the flexibility to choose how we want to handle this would be welcome!