Degree
Master of Science in Data Science
Department
Department of Computer Science
Faculty/ School
School of Mathematics and Computer Science (SMCS)
Date of Submission
Fall 2024
Supervisor
Khawaja Abdul Hafeez, Visiting Faculty, Department of Computer Science
Keywords
SQL Query Generation, NLP, OpenAI
Abstract
InsightsDesk is a simple chatbot that works as a helpdesk, providing technology and database insights in English to non technology, non technical C-level executives. The chatbot uses a Retriever augmented generation (RAG) approach, while leveraging Tavily API to handle non DB related questions and Open AI API for DB related queries and queries generation. Then, when a query is identified as being database related, the system transforms the question into an SQL query using a query generator prompt. This interpreter prompt interprets the results into user friendly responses that are then executed against a PostgreSQL database successfully. InsightsDesk responds with generic conversation abilities similar to most ChatGPT model in case the query is not tied to the data base, it provides solution for literally any question. With this dual capability users get clean, easy to understand answers and can make informed decisions without having to know a lot about technology. InsightsDesk not only makes data accessible for the executive decisions, but it’s also an example of how advanced AI technologies are bridging the gap between the complexity of data management systems and the operations of strategic business systems.
Document Type
Restricted Access
Submission Type
Research Project
Recommended Citation
Ahmed, N. (2024). InsightsDesk: An LLM-Powered SQL Agent for Executive Data Insights (Unpublished graduate research project). Institute of Business Administration, Pakistan. Retrieved from https://ir.iba.edu.pk/research-projects-msds/35
