Querying CSV and Parquet exports inside a SQL IDE with DuckDB

Description and use

Use omni-sql to import CSV/Parquet and run SQL queries locally for quick data analysis without database setup.

Problem

Investigate a CSV export for duplicates, grouping, or comparison without lengthy database setup.

Outcome

The joined query revealed that South reached 84% of its target, demonstrating the ability to combine external CSV data with database query results locally.

Workflow

• Run the sales query on PostgreSQL and export the full result to a file.
• Import targets.csv into the local DuckDB workspace using omni-sql.
• Run a JOIN query in the local SQL tab to combine the query result with the CSV data.
• Review the output to identify gaps and target percentages.

Evidence

Someone sends you a CSV export and asks you to investigate it. You need to check duplicates, group records, or compare it with another export.

Source

You can import CSV or Parquet and query it with SQL inside the IDE.

Source

The demo uses fictional sales rows from PostgreSQL and a CSV of regional targets.

Source

The result shows South reaching 84% of its target.

Source

The app runs locally and doesn't require an account.

Source