Exploring Data

After ingesting data, you land on the Explorer — the main workspace for viewing, querying, and understanding your data.

Explorer layout

The Explorer has three panels:

1. Left sidebar — file tree showing all ingested tables. Click a table to select it.
2. Center panel — data grid with column headers, sortable columns, and inline profiling. A tabbed SQL workspace lives below (or can be toggled).
3. Right panel — AI widget recommendations. Charts stream in automatically after ingestion.

File tree sidebar

The sidebar lists every ingested table, named after the source file. You can:

• Click a table to load it in the data grid
• Unmount a table by clicking the X button next to the file size — this drops the DuckDB table and frees memory
• See file sizes and row counts at a glance

When you upload a folder, files appear in a nested tree structure matching the original directory layout.

Data grid

The data grid shows a paginated, scrollable view of the selected table's data. Features:

• Column headers — click to see data type, null count, and value distribution
• Sorting — click a column header to sort ascending/descending
• Row count — displayed in the footer
• Horizontal scroll — tables with many columns scroll horizontally

The grid displays the first rows of the table. For full data exploration, use the SQL workspace.

SQL workspace

The SQL workspace is a code editor where you can write and run DuckDB SQL queries against your tables.

-- Example: top 10 categories by revenue
SELECT category, SUM(amount) AS total_revenue
FROM t_sales
GROUP BY category
ORDER BY total_revenue DESC
LIMIT 10

Key features:

• Syntax highlighting — DuckDB SQL dialect
• Auto-complete — table and column names
• Results grid — query results appear below the editor
• Query tabs — run multiple queries in separate tabs

The workspace auto-initializes with SELECT * FROM <table> LIMIT 100 when you select a table.

Column profiling

Click any column header in the data grid to see a quick profile:

• Data type — VARCHAR, INTEGER, DOUBLE, TIMESTAMP, etc.
• Null count — how many rows have null values
• Unique values — count of distinct values
• Distribution — a mini histogram or value frequency chart

This helps you understand your data's shape before running queries or generating widgets.