This project involves developing an AI-powered Text-to-SQL chatbot that enables users to interact with a relational database using natural language. The system interprets user questions, generates corresponding SQL queries, executes the queries against a MySQL database, and presents the retrieved information in a user-friendly format.
The implementation begins with designing the system architecture and database schema, including the preparation of an Entity-Relationship diagram. Structured data is then loaded into MySQL to establish the database environment required for query execution. The chatbot is implemented to interpret natural language inputs and generate SQL statements based on the user's intent and the available database structure.
The project also explores the evaluation of AI-generated outputs using RAGAS or other suitable evaluation techniques, depending on the implemented workflow. This helps assess the quality of generated responses and identify areas for improvement. The final stages focus on future enhancements, including improving query accuracy, handling complex database relationships, and strengthening the reliability of generated SQL.
Overall, the project demonstrates how Generative AI and natural language processing can simplify database interaction by allowing users to retrieve structured information without manually writing SQL queries.