Talk to a large collection of public data

It is a query interface that uses natural language so anyone can explore the Data Basis public datalake. It uses Google's Gemini model to turn your questions into precise data queries.

No. The Chatbot was designed so anyone can explore public data using natural language. Even extracting the data used in an answer requires no prior SQL knowledge: the chatbot generates the query and lets you go straight to the data you need.

The chatbot is connected to more than 740 tables that make up DB's public datalake. It is a valuable and constantly expanding source, with data on diverse topics: economy, health, education, demography, environment and much more.

Yes. In addition to the text analysis, you receive the code to download the data via BigQuery. This gives you the flexibility to take the information into Excel, Google Sheets or visualization tools and create your own maps and dashboards.

Yes. We offer a 7-day free trial so you can get to know the tool. After that period, the chatbot is part of a subscription plan that guarantees unlimited questions and helps us keep our public datalake free.

Every chatbot analysis comes with an SQL query. You can copy that code and use it in BigQuery (a Google tool) to extract the full table and create your own visualizations or reports.

Yes. Subscribers get priority support via email and Discord. You can also send questions and suggestions directly to our community and technical team.

The Chatbot always returns an SQL query along with the answer. You can copy that query and use Google BigQuery — Google's data query and visualization tool — to extract a table with the data used in the analysis.

To do so, follow these steps:

1. Go to the Google BigQuery site.
2. If you land on the product's home page, click something like Try in Console or Console to open the query interface.
3. Sign in with a Google account and create a project, if you don't have one yet.
4. Once the project is created, open SQL Query (or the equivalent in the console) to access the editor.
5. Paste the query provided by the chatbot and run it.
6. After running it, you can save the results in your preferred format or open them with other Google tools for visualization.

If you run into trouble with this flow, our team and community can help on the Data Basis Discord.

For complex questions, we recommend breaking the request into smaller, simpler questions. Another useful tip is to mention the specific name of a column (such as "municipality") or explicitly ask the chatbot to search within a dataset you already know.

Beta means the chatbot is constantly improving. That's why your feedback through the like and dislike buttons (👍/👎) on each answer is essential for us to calibrate the model and improve the accuracy of the results.