Best way to minimize latency when using large data sets

I use the Toucan in a multi-tenant situation. I sometimes need to pull in large datasets that have thousands of lines based on duration periods. I'm trying to decide the best way to do this... If I use a live query, it can take too long to fetch and render the data in Toucan. Using the pre-fetched data, that dataset becomes very large and also creates delays in the rendering to the screen. Is there a way to construct a date function with the ability to pull data over a week, or two weeks, allowing the size of the dataset to remain manageable? Without the date picker native functionality, the user might can pick an entire month and have a very long refresh event. Has anyone else solved this problem?