Sajjadur Rahman

Building, Serving, and Growing a Conversational AI Assistant for Enterprise

Sumit Bhatia, Vidit Bhatia, Uttaran Bhattacharya, Victor S. Bursztyn et autres

Enterprises increasingly seek conversational AI systems that can reason over both structured and unstructured knowledge to enhance productivity and decision-making. Building such systems requires integrating heterogeneous data, grounding responses in domain-specific context, ensuring compliance and explainability, and maintaining quality through continual improvement. …

us (code pays fourni par la source)

From Facts to Folklore: Evaluating Large Language Models on Bengali Cultural Knowledge

Association for Computational Linguistics 2025, Anika Ahmed, Nafis Chowdhury, Md. Moinul Haque et autres

Recent progress in NLP research has demonstrated remarkable capabilities of large language models (LLMs) across a wide range of tasks. While recent multilingual benchmarks have advanced cultural evaluation for LLMs, critical gaps remain in capturing the nuances of low-resource cultures. Our work …

bd (code pays fourni par la source)

From Facts to Folklore: Evaluating Large Language Models on Bengali Cultural Knowledge

Mst. Mafruha Haque, Nazia Tasnim, Md. Istiak Hossain Shihab, Sajjadur Rahman et autres

Recent progress in NLP research has demonstrated remarkable capabilities of large language models (LLMs) across a wide range of tasks. While recent multilingual benchmarks have advanced cultural evaluation for LLMs, critical gaps remain in capturing the nuances of low-resource cultures. Our work …

Evaluation and incident prevention in an enterprise AI assistant

Akash V. Maharaj, David T. Arbour, Daniel C. Lee, Uttaran Bhattacharya et autres

Abstract Enterprise AI Assistants are increasingly deployed in domains where accuracy is paramount, making each erroneous output a potentially significant incident. This paper presents a comprehensive framework for monitoring, benchmarking, and continuously improving such complex, multi‐component systems under active development by multiple …

us (code pays fourni par la source)

Towards Probabilistic Question Answering Over Tabular Data

Chen Shen, Sajjadur Rahman, Estevam Rafael Hruschka Junior

Current approaches for question answering (QA) over tabular data, such as NL2SQL systems, perform well for factual questions where answers are directly retrieved from tables. However, they fall short on probabilistic questions requiring reasoning under uncertainty. In this paper, we introduce a …

LakeVisage: Towards Scalable, Flexible and Interactive Visualization Recommendation for Data Discovery over Data Lakes

Yihao Hu, Jin Wang, Sajjadur Rahman

Data discovery from data lakes is an essential application in modern data science. While many previous studies focused on improving the efficiency and effectiveness of data discovery, little attention has been paid to the usability of such applications. In particular, exploring data …

us (code pays fourni par la source)

Effectiveness of Prompt Optimization in NL2SQL Systems

Sairam Gurajada, Eser Kandogan, Sajjadur Rahman

NL2SQL approaches have greatly benefited from the impressive capabilities of large language models (LLMs). In particular, bootstrapping an NL2SQL system for a specific domain can be as simple as instructing an LLM with sufficient contextual information, such as schema details and translation …

Enhanced Detection of Q and S Peaks in ECG Signals with Noise Resilience

Sajjadur Rahman

This study introduces an improved methodology for detecting $\mathbf{Q}$ and $\mathbf{S}$ peaks in electrocardiogram (ECG) signals, addressing challenges in accuracy and resilience under noisy conditions. Building on the classical Pan-Tompkins algorithm, the proposed approach incorporates enhanced preprocessing steps, including modified filtering using …

bd (code pays fourni par la source)

Towards Operationalizing Heterogeneous Data Discovery

Jin Wang, Yanlin Feng, Chen Shen, Sajjadur Rahman et autres

Querying and exploring massive collections of data sources, such as data lakes, has been an essential research topic in the database community. Although many efforts have been paid in the field of data discovery and data integration in data lakes, they mainly …

MageSQL: Enhancing In-context Learning for Text-to-SQL Applications with Large Language Models

Chen Shen, Jin Wang, Sajjadur Rahman, Eser Kandogan

The text-to-SQL problem aims to translate natural language questions into SQL statements to ease the interaction between database systems and end users. Recently, Large Language Models (LLMs) have exhibited impressive capabilities in a variety of tasks, including text-to-SQL. While prior works have …

LakeVisage: Towards Scalable, Flexible and Interactive Visualization Recommendation for Data Discovery over Data Lakes

Yihao Hu, Jin Wang, Sajjadur Rahman

Data discovery from data lakes is an essential application in modern data science. While many previous studies focused on improving the efficiency and effectiveness of data discovery, little attention has been paid to the usability of such applications. In particular, exploring data …

FactLens: Benchmarking Fine-Grained Fact Verification

Kushan Mitra, Dan Zhang, Sajjadur Rahman, Estevam Rafael Hruschka Junior

Large Language Models (LLMs) have shown impressive capability in language generation and understanding, but their tendency to hallucinate and produce factually incorrect information remains a key limitation.To verify LLM-generated contents and claims from other sources, traditional verification approaches often rely on holistic …

us (code pays fourni par la source)