Open Access Open Access  Restricted Access Subscription Access

NulAware AI: A Retrieval-Augmented Natural Language Assistant for Intelligent Data Profiling

Ms.R.SHIVA DHARSHINI, SRIDEVI SV, SANTHIYA K, YALINI B, SUBIKSHA K

Abstract


Understanding the structure and quality of a dataset is the first and most time-consuming step of any data project, and it normally demands programming skill. This paper presents NulAware AI, a conversational data profiling assistant that lets users upload a CSV file, explore it through an interactive dashboard, generate a profiling report and ask questions in plain English. The system computes the dataset profile (schema, missing values, summary statistics and correlations) with Pandas and ydata-profiling, converts it into text chunks, embeds them with a HuggingFace model and stores them in ChromaDB. A LangGraph agent routes each question, retrieves the most relevant chunks and supplies them as context to Gemini 2.5 Flash, which produces a grounded answer. Because numerical facts are computed by proven libraries and the language model only explains retrieved context, the approach reduces hallucination while keeping the tool accessible to non-programmers. The interface is built with Streamlit and the services with FastAPI.


Full Text:

PDF

References


Rahm E., Do H.H. Data Cleaning: Problems and Current Approaches. IEEE Data Engineering Bulletin. 2000, 23(4), 3-13p.

Kandel S., Paepcke A., Hellerstein J.M., Heer J. Enterprise Data Analysis and Visualization: An Interview Study. IEEE Transactions on Visualization and Computer Graphics. 2012, 18(12), 2917-2926p.

Vaswani A. et al. Attention Is All You Need. Advances in Neural Information Processing Systems. 2017.

Lewis P. et al. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. Advances in Neural Information Processing Systems. 2020.

Karpukhin V. et al. Dense Passage Retrieval for Open-Domain Question Answering. Proceedings of EMNLP. 2020.

Reimers N., Gurevych I. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks.

Proceedings of EMNLP-IJCNLP. 2019.

Yao S. et al. ReAct: Synergizing Reasoning and Acting in Language Models. Proceedings of ICLR. 2023.

Gemini Team, Google. Gemini: A Family of Highly Capable Multimodal Models. arXiv:2312.11805. 2023.

Yu T. et al. Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task. Proceedings of EMNLP. 2018.

YData. ydata-profiling Documentation. https://docs.profiling.ydata.ai/


Refbacks

  • There are currently no refbacks.