A Quick Start Guide to Exploring Your CSV Data with SQL

Whether you are an experienced data analyst or just need to grab a few metrics from a sales report, CSV GPT is designed to eliminate friction. Here is how you can go from a raw CSV file to derived insights in seconds.

1. Upload Your Data

Navigate to the app and simply drag and drop your CSV file onto the screen. The internal DuckDB WASM engine will instantly load your data and generate a clear preview of the schema structure and row count, right inside your browser.

2. Manual SQL Mode for Pros

Sometimes you know exactly what you want and it's faster to write the query yourself. You can toggle our Manual SQL Mode to bypass the AI completely.

• Type standard SQL directly into the input box.
• Your table is referenced as my_data.
• Example: SELECT status, COUNT(*) FROM my_data GROUP BY status;

Because processing is handled via DuckDB locally, query execution is incredibly fast, even for millions of rows.

3. AI-Assisted Exploration

If you prefer a conversational approach, ensure your LLM is configured (either an API key or a local setup). Simply ask normal questions like:

"What are the top 5 regions by total sales volume last quarter?"

The AI will understand the context of your data schema, formulate the perfect DuckDB SQL query, and run it to fetch the results. It will even summarize the final output for you or generate insightful visualizations.

It's that simple. Powerful data exploration with zero infrastructure.