DATA VOICE AI: A Conversational Business Intelligence Assistant Using RAG

The presented work demonstrates the development of a conversational Business Intelligence platform capable of simplifying analytical interaction with organizational databases using natural language queries. The system integrates Retrieval-Augmented Generation, Natural Language Processing, semantic retrieval, SQL generation, and dynamic visualization within a unified web-based environment. The developed platform reduces dependency on manual SQL writing and improves accessibility for users without technical database knowledge.   The implemented architecture successfully supports analytical operations across Sales, Finance, Human Resources, and Inventory databases through conversational interaction. Context-aware retrieval using FAISS similarity search improved query interpretation during SQL generation, while schema-aware prompting enhanced analytical processing efficiency. The generated SQL statements were executed successfully on SQLite databases, and the retrieved results were presented through KPI cards, analytical tables, and graphical visualizations. The platform also demonstrated voice-enabled analytical querying, multi-database interaction, and export functionality for analytical reports. The obtained implementation results confirmed that conversational AI and Retrieval-Augmented Generation can effectively improve Business Intelligence interaction and simplify organizational analytical workflows. The developed system provides a practical foundation for future AI-driven conversational analytical platforms and enterprise Business Intelligence applications.

[1] Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin, “Attention Is All You Need,” Advances in Neural Information Processing Systems (NeurIPS), Vol. 30, 2017.   [2] Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al., “Language Models are Few-Shot Learners,” Advances in Neural Information Processing Systems (NeurIPS), Vol. 33, 2020, pp. 1877–1901.   [3] Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems (NeurIPS), Vol. 33, 2020, pp. 9459–9474.   [4] Victor Zhong, Caiming Xiong and Richard Socher, “Seq2SQL: Generating Structured Queries from Natural Language Using Reinforcement Learning,” arXiv preprint arXiv:1709.00103, 2017.   [5] Bailin Wang, Richard Shin, Xiaodong Liu, Oleksandr Polozov and Matthew Richardson, “RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers,” Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 2020, pp. 7567–7578.   [6] Jeff Johnson, Matthijs Douze and Hervé Jégou, “Billion-scale Similarity Search with GPUs,” IEEE Transactions on Big Data, Vol. 7, No. 3, 2021, pp. 535–547.   [7] Jacob Devlin, Ming-Wei Chang, Kenton Lee and Kristina Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” Proceedings of NAACL-HLT, 2019, pp. 4171–4186.   [8] Nils Reimers and Iryna Gurevych, “Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks,” Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing, 2019, pp. 3982–3992.   [9] Sebastián Ramírez, “FastAPI Framework Documentation,” FastAPI Official Documentation, 2024.   [10] SQLite Development Team, “SQLite Documentation,” SQLite Official Documentation, 2024.