In your CSV action you can choose the source, meaning the data you’re to going to export as CSV. Even if you have filters applied, they are only applied in the front end, the entire data set will be downloaded as a CSV.
You would need, in your glide table, to add a query column, and filter only the desired element, the in the CSV action, use your query as source. Then the CSV will only have data from the query.
If it’s not a static export, meaning your criteria for the export change in time, you will need to use a custom filter. Robert Petitto made some great videos about it, Ill see if I can find them.
It’s most likely not a “static export” per Lucas’ definition above, since you allow users to filter them on the front end. You can get started with this.
In-app filtering only applies to the collection itself on the front end. You can’t use it to filter anything in the back end for export. You would have to build your own custom filters.
As long as you build your custom filter correctly, you will be able to generate a CSV through a query with those filters, and allow the user to download that file.
The Generate CSV integration isn’t supported with Big Tables, but there are other approaches that can be used. Three separate methods that I’ve had success with:
The traditional joined list via a relation/query. Works okay but could be limited to 100 rows depending on any computed column types involved.
A joined list of joined lists, then use JavaScript to generate the CSV.
Use the Glide API to run a query from Make, then generate the CSV on the Make side. This is a pretty robust method, but can only be used if you are filtering on non-computed columns. I helped Jack make a video that describes this method, see below: