How to speed up several topN queries when using Pivot on top of Broker node

Hi Druid team,
My team is working on Pivot with Druid as data store :slight_smile: We have used production cluster recommended settings for all Druid nodes. We are ingesting data at the rate of 3billion records/day. Pivot is used on top of Druid for exploration. It gives quick results when we use basic query. However, the query times out when we try 5-6 levels of topN queries on various dimensions. We have caching enabled at broker and historical node. Any suggestion of tuning any parameter in order to fasten query result when using various topN queries. Any help is appreciated.