EFF BI: Effortless Business Intelligence -Text-to-Vis from LLM Agents
Traditional Business Intelligence (BI) tools rely on complex user interfaces and modeling languages, creating significant accessibility barriers for non-technical users.This often forces domain experts to rely on BI developers to translate and refine their requirements, resulting in time-intensive workflows.In this paper, we introduce Eff BI (Effortless Business Intelligence), an open-source, LLM-powered Text-to-Visualization (Text-to-Vis) tool that enables users to retrieve data from enterprise databases, and generate graphs and charts instantly through natural language.Eff BI's open-source architecture supports plug-and-play functionality with proprietary LLMs, ensuring data remains secure within company servers.Our application is fully open-source and accompanied by documentation for both users and future developers.An example of the visualization and corresponding SQL output is shown in Figure 1.: https://github.com/Vshnv2001/eff-bi: https://eff-bi.vercel.app/
Authors
- Tan Wee Kian Justin
- Vinitha Muralidharan
- H Bryan
- Muhammad Reyaaz
- Suzanna Sia
Institutions
- National University of Singapore (SG)
Publication Details
- Published
- 2025-05-08
- DOI
- https://doi.org/10.1145/3701716.3715173
- Primary Topic
- Semantic Web and Ontologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00