Working with Large Datasets & Results

Guidance for working with large datasets and exporting results.

Exporting & Saving Results

Export Type	Description	Examples
Quick Exports	Export analysis results for use elsewhere. The export includes all rows from the result, not just what’s displayed in the preview.	• “Export this as CSV” • “Save these results as an Excel file”
Saving to Workspace	For results you’ll reference again, save to your workspace. You can add descriptions and tags to make it findable later.	• “Save this analysis to my Reports folder”
Creating Derived Tables	Save analysis results as queryable datasets. You can then ask questions against this derived table without re-running the original analysis.	• “Save this result as a new table”

Zark handles large datasets efficiently, but some awareness of scale helps you work effectively.

Dataset Size	Performance	Notes
Thousands of rows	Process instantly
Hundreds of thousands of rows	Still fast
Millions of rows	Require more processing time but work fine
Tens of millions of rows	Supported but benefit from filtering early	Progress indicators show processing status. You don’t need to wait for full processing before asking questions—once enough data is loaded, Zark can start responding.

For very large datasets, specific questions perform better than open-ended exploration:

“Show me all records” on a 50-million-row table will take time
“Show me the top 100 customers by revenue in Q4” filters and limits, returning quickly

Start with filtered subsets or aggregations rather than raw records:

Zark applies sensible limits automatically. You can be explicit: