Explorer

Explorer is a powerful workspace that allows you to ask questions in natural language and automatically generate SQL queries against your connected datasets. It’s designed to make exploring data fast, intuitive, and accessible — whether you’re an analyst, developer, or domain expert.

How Explorer Works

1. Ask a Question: Type your question in plain language (e.g., “What were total sales by region last quarter?”).
2. Automatic SQL Generation: Explorer converts your question into an SQL query that runs against the selected dataset.
3. Review the Results: Instantly see both the generated SQL and the query output.
4. Edit as Needed: If you need to refine the logic, filters, or structure, you can directly modify the SQL.
5. Save as a K-Shot: If the question–query pair is useful for future learning, save it as a K-Shot to guide the model on similar queries going forward.

Auto-Generated K-Shots

Explorer can also automatically generate K-Shots based on your dataset and schema. This gives you a fast starting point for building a high-quality library of training examples for SQLGENAI.

• One-Click Generation: Create multiple example question–SQL pairs in seconds.
• Editable: Review and refine the generated SQL directly in Explorer.
• Curated Quality: Use the generated examples as-is or adjust them to match your business logic.

Why K-Shots Matter

K-Shots are example question–SQL pairs that help the system learn how to respond accurately to similar queries in the future. High-quality K-Shots:

• Improve the accuracy and relevance of generated SQL.
• Provide better consistency across different users and questions.
• Reduce the need for repeated manual query writing.

Why Explorer Is Powerful

• Speed: Move from a question to a working query in seconds.
• Flexibility: Edit or extend the SQL as needed to meet your specific needs.
• Scalability: Build a robust foundation of examples without starting from scratch.
• Collaboration: Share K-Shots across your team to standardize how common questions are answered.

Tip: Reviewing and refining your K-Shots is key to getting the best results from SQLGENAI. Even auto-generated examples benefit from a quick human check.