How to filter out unnecessary field from the kibana output

Hi, I am wondering if we can write queries that automatically removes the fields in the output that we don't want. I am asking this because my nested aggregation keeps crashing my cluster. If we could remove millions of fields that are not useful to us , we can save space and prevent excessive memory use. For example, in the figure below, how to get just the fields that has been highlighted. Is there any way?
QUERY_link