AI-Powered SQL Query Generation

Target Audience

• Data Analysts: Individuals analyzing database schemas and querying data efficiently.
• Developers: Those who need to automate SQL query generation and execution without manual intervention.
• Data Scientists: Professionals who require quick access to structured data for analysis.
• Business Intelligence Professionals: Users who want to extract insights from databases using natural language queries.
• Educators and Students: Anyone learning about database interactions and automation workflows.

Problem Solved

This workflow addresses the challenge of generating SQL queries from a database schema based on user input. It automates the process of:

• Extracting Database Schema: Users can retrieve the structure of the database without needing to know SQL.
• Generating Queries: It allows users to formulate queries using natural language, making it accessible to those without SQL expertise.
• Executing Queries: Users can run their generated queries and receive formatted results, streamlining data retrieval and analysis.

Workflow Steps

1. Trigger: The workflow is manually initiated via a trigger node.
2. List Tables: It retrieves all tables from the specified MySQL database.
3. Extract Schema: For each table, the schema is extracted and stored for further processing.
4. Convert to JSON: The schema data is converted to a binary JSON format for efficient storage.
5. Save Locally: The binary JSON schema is saved as chinook_mysql.json locally.
6. Chat Interaction: Users can interact with the system through a chat interface where they input their queries.
7. Load Schema: Upon receiving a query, the workflow loads the schema from the local file to ensure quick access.
8. Generate SQL Query: The AI agent generates an SQL query based on the user input and the schema.
9. Check Query: It checks if the generated query is valid and exists.
10. Run SQL Query: If a valid query exists, it executes the query against the database.
11. Format Results: The results from the query are formatted for readability and combined with the chat response.
12. Output: Finally, the combined output of the chat response and the query results is provided to the user.