Query

The Query cell is your fully featured SQL editor inside Livedocs.
It lets you query data from either:

• A data warehouse (Postgres, Snowflake, BigQuery, etc.)
• A file (CSV, Parquet, etc., queried with DuckDB)

Selecting a Source

• Pick your source from the dropdown.
• SQL dialect automatically matches the source:
  • Postgres connector → Postgres SQL
  • Snowflake connector → Snowflake SQL
  • Clickhouse connector → Clickhouse SQL
  • BigQuery connector → BigQuery SQL
  • Files → DuckDB SQL (default)
• Note: Postgres tables are attached in-memory to DuckDB for queries.

Execution & Performance

• Queries run in the warehouse compute where possible.
• Results are returned as a Polars DataFrame.
• Caching:
  • On the first run, the DataFrame is cached in memory.
  • Subsequent runs (without query changes) return from cache instantly.
  • To force fresh results, click the dropdown on the Run button → Run without cache.
• Results are paginated for large datasets.

Editor Features

• Fullscreen Mode: Hover over the cell header → click the three dots (More Menu) → Expand.
• AI Assistance: Click the bolt icon ⚡ to open a new chat with your preferred model to:
  • Write queries
  • Debug queries
  • Explain results

Working with Results

• Click result column headers to:
  • Filter
  • Sort
  • Color cells
• Add column-level aggregate functions like:
  • SUM
  • COUNT
  • AVG

Example

-- Querying a local CSV with DuckDB
SELECT customer_id, SUM(total) AS total_spent
FROM 'sales.csv'
GROUP BY customer_id
ORDER BY total_spent DESC
LIMIT 10;