Project overview
Bhavya Pandey, Ankush Maheshwari
This research aims to improve the existing Text-to-SQL models to help organizations achieve more efficient and accessible database interactions. The goal is to enable companies to generate highly optimized SQL queries from natural language inputs, reducing the need for extensive technical expertise in SQL syntax and allow organizations to streamline their data querying processes, potentially leading to significant cost savings. We plan to leverage SPIDER/BIRD datasets along with advanced prompting and RAG/TAG techniques with LLMs like Llama and GPT to improve SQL query accuracy & efficiency.
Highlight of the projects
To be declared.