BigQuery Text-to-SQL Agent
Orchestrated by Mochammad Reyhan Mauluddi.
Ask questions about e-commerce data in English or Indonesian; the agent handles the SQL.
Query e-commerce data without writing SQL
This agent accepts natural-language questions about e-commerce data, translates them into SQL, runs the query against BigQuery, and returns the answer as a table, a chart, and a written summary. The SQL behind every answer is always shown, so results can be independently verified.
Dataset: Google BigQuery’s public thelook_ecommerce dataset — 7 tables (users, orders, order items, products, inventory items, distribution centers, and events) totaling approximately 3.35 million rows. Coverage verified current as of July 16, 2026.
How to use it
- 1
Ask your question
English or Indonesian is fine — no SQL knowledge required.
- 2
The agent takes it from there
Your question is translated into SQL, checked for safety (read-only, restricted to whitelisted tables), and executed against BigQuery.
- 3
Review the result from every angle
The SQL query, the result table, a chart, a diagram of the agent's workflow, and a narrative summary — open whichever tab you need.
- 4
Export when you're done
Download the table as CSV or XLSX, or the chart as a JPG.