eSapiens: A Real-World NLP Framework for Multimodal Document Understanding and Enterprise Knowledge Processing

TL;DR

eSapiens is a comprehensive enterprise question-answering framework that integrates structured and unstructured data sources, utilizing advanced retrieval and generation techniques to improve answer accuracy and grounding in real-world scenarios.

Contribution

The paper introduces eSapiens, a novel unified system combining Text-to-SQL and hybrid RAG modules with citation verification for enterprise document understanding.

Findings

eSapiens outperforms baseline models in relevance and quality.

The system effectively reduces hallucinations and improves grounding.

Optional strict grounding enhances reliability in critical applications.

Abstract

We introduce eSapiens, a unified question-answering system designed for enterprise settings, which bridges structured databases and unstructured textual corpora via a dual-module architecture. The system combines a Text-to-SQL planner with a hybrid Retrieval-Augmented Generation (RAG) pipeline, enabling natural language access to both relational data and free-form documents. To enhance answer faithfulness, the RAG module integrates dense and sparse retrieval, commercial reranking, and a citation verification loop that ensures grounding consistency. We evaluate eSapiens on the RAGTruth benchmark across five leading large language models (LLMs), analyzing performance across key dimensions such as completeness, hallucination, and context utilization. Results demonstrate that eSapiens outperforms a FAISS baseline in contextual relevance and generation quality, with optional strict-grounding controls for high-stakes scenarios. This work provides a deployable framework for robust, citation-aware question answering in real-world enterprise applications.