Aug 2026· International Journal for Research in Applied Science and Engineering Technology· Vol 14, pp. 1094-1100· 0 citations
TL;DR
An Intelligent Document Processing Platform using Retrieval-Augmented Generation to enable accurate, context-aware, and reliable document intelligence and offers a practical and scalable framework for intelligent document understanding, semantic search, and AI-assisted question answering in modern knowledge management environments.
Abstract
The rapid growth of digital documents across enterprises, educational institutions, healthcare organizations, and
research environments has created significant challenges in information retrieval and knowledge management. Traditional
keyword-based search systems often fail to capture the semantic meaning and contextual relationships present in large
collections of unstructured documents, resulting in inefficient information access and reduced productivity. This study presents
an Intelligent Document Processing Platform using Retrieval-Augmented Generation (RAG) to enable accurate, context-aware,
and reliable document intelligence. The proposed system integrates document ingestion, text extraction, preprocessing,
chunking, embedding generation, vector-based semantic retrieval, and Large Language Models (LLMs) within a unified
framework. Uploaded documents are transformed into vector embeddings and stored in a vector database, enabling efficient
similarity-based retrieval. When a user submits a natural language query, the system retrieves the most relevant document
segments and augments them as contextual knowledge for response generation. This retrieval-grounded approach reduces
hallucinations, improves factual accuracy, and enhances the relevance of generated answers. The platform is implemented using
Python, Streamlit, LangChain, embedding models, and vector databases to provide an interactive and scalable solution for
document-centric knowledge discovery. Experimental evaluation demonstrates improved retrieval effectiveness, faster access to
relevant information, and enhanced user experience compared with conventional document search approaches. The proposed
system offers a practical and scalable framework for intelligent document understanding, semantic search, and AI-assisted
question answering in modern knowledge management environments.
Artificial Intelligence (AI) and Large Language Models (LLMs) have significantly transformed knowledge management by enabling intelligent, context-aware, and automated information access. However, standalone LLMs often suffer from limitations such as outdated knowledge, hallucinated responses, lack of domain-specific expertise, and limited transparency, reducing their reliability in enterprise and research applications. Retrieval-Augmented Generation (RAG) has emerged as an effective solution by combining language models with external knowledge retrieval, allowing responses to be generated using up-to-date and relevant information. This study proposes a comprehensive Retrieval-Augmented Generation framework for intelligent knowledge management systems. The framework integrates document acquisition, preprocessing, semantic embedding generation, vector database indexing, document retrieval, prompt augmentation, LLM-based response generation, response validation, and continuous knowledge base updates. It supports diverse knowledge sources, including enterprise databases, technical documents, digital libraries, and research repositories, while incorporating sparse, dense, hybrid retrieval, and neural reranking techniques to improve retrieval accuracy. The proposed framework is evaluated using retrieval precision, recall, F1-score, response relevance, latency, grounding accuracy, and user satisfaction. Results demonstrate improved semantic understanding, reduced hallucinations, enhanced factual correctness, and real-time knowledge updates compared with conventional keyword-based knowledge management systems. The study also discusses future directions, including multimodal RAG, graph-enhanced retrieval, federated knowledge management, continual learning, and autonomous enterprise knowledge assistants, establishing RAG as a robust foundation for trustworthy and intelligent knowledge-driven AI systems.
Louis Pouzin, J. Arsac· International Journal of Mod...· 0 citations
Due to the surge in digital document creation, intelligent systems that can effectively retrieve accurate information from unstructured PDF content have become a necessity. Current keyword-based search techniques lack semantic meaning, while large language models (LLMs), when utilized individually, produce hallucinated results when they are not contextually aware of document content. In this paper, we introduce a novel real-time question answering system for PDF content using a lightweight approach to Retrieval-Augmented Generation (RAG). Our system combines a HuggingFace sentence transformer model for semantic embeddings with a FAISS vector similarity search for efficient retrieval, leveraging a Groq-hosted LLaMA 3.3 70B model for context-grounded answer generation. The system is built using LangChain for orchestration and Streamlit for web-based deployment, and requires no GPU, no fine-tuning, and no pre-indexed corpus. Evaluation across five complementary metrics—Exact Match, Token-level F1, BLEU, ROUGE, and semantic similarity—demonstrates strong alignment between generated and reference answers, with a peak semantic similarity of 0.8709 and a mean Token F1 of 0.4701, confirming the viability of this approach for real-world document intelligence.
Mukesh Lakshmi Sai Medikonda, Kalva Vishnu Teja, K. Greeshma et al.· International Conference Com...· 0 citations
Retrieval-Augmented Generation (RAG) has established itself as a compelling strategy for grounding large language model outputs in documentary evidence. However, production deployments continue to rely almost exclusively on homogeneous text corpora, even as enterprise repositories grow increasingly heterogeneous blending technical schematics, radiological images, annotated diagrams, and unstructured prose within the same archival system. This mismatch between system design and data reality motivates the present work. We propose a Multimodal RAG framework that unifies text and image retrieval through four tightly coupled components: a dual-stream embedding engine, a learned four-class AI query router, an adaptive confidence threshold, and a session-aware context store. Text is encoded with the allmpnet-base-v2 Sentence Transformer; images are embedded in the same 768-dimensional space via a domain-adapted CLIP ViT-L/14 model fine-tuned on approximately 120,000 technical and clinical text-image pairs. Concatenating 768-d vectors yields a 1,536-d composite query that drives a single approximate nearest-neighbor (ANN) search simultaneously across both modalities. A fine-tuned DistilBERT router assigns each query to one of four retrieval pathways text-only, image-only, hybrid, or conversational at 91.3% accuracy. Experiments were run on three enterprise corpora (engineering manuals, clinical case summaries, and legal paperwork) with the results being a mean F1@5 of 0.90, an increase of 22% over a dense text-only baseline, and a reduction of 49% in the number of hallucinations. The median first-token latency is 1.34s, which meets the interactive-use target deployment-contexts. These results show that it is possible to implement modality-aware retrieval in a simple and realistic setting with real-world organizational constraints and that the results are reproducible.
E.Vijayakumar, Ganesh A· 2026 4th International Confe...· 0 citations
Extraction of information from the corpus is an complicated task considering the information overload, format diversity, multilingually, and multimodal behavior of the information. The proposed work mainly targets accurate information extraction from South Indian Language documents, such as Malayalam, Kannada, Telugu, and Tamil, using the Retrieved Augmented Generation [RAG] model. The main problem associated with low-resource languages such as these is their dense morphology and agglutinative behavior. The goal of this research work is to incorporate RAG with language models to avoid hallucinated responses and successfully perform relevant information extraction from multilingual documents. In this work, RAG will be incorporated with leading FM models and versions of Gemini, Lamma, and DeepSeek. Implementation achieved using the Google Gemini API, Qdrant vector database, advanced Multilingual embedding and efficient prompt engineering. The application allows users to upload a corpus in their specific language and ask any question; the model will retrieve the accurate answer based on the query. The implementation details are publicly available and can be verified through the [link-https://github.com/siv3sh/Multilingual_document_Reader. git] GitHub repository. Each model varies in performance in terms of Latency, long-term dependencies, supporting tokens and accuracy in multilingual completions.
Gobi Ramasamy, Nisha Elsa Varghese, Antoine B. Bagula et al.· 2026 ITU Kaleidoscope - AI a...· 0 citations
Managing personal documents remains a hassle: files accumulate across cloud drives, email, and local devices in formats that mix scanned images with digital text, and conventional retrieval based on filenames and folder hierarchies handles neither ambiguous queries nor cross-format access well. This paper presents Intelli-Docs, a personal document assistant that combines retrieval- augmented generation (RAG) with multimodal retrieval so that users can locate and question their documents through natural language queries. The system ingests documents through three coordinated pipelines. A text pipeline extracts and chunks document text, embeds the chunks, and stores them in a vector index. A query pipeline classifies each request as text retrieval, image retrieval, or general conversation and routes it accordingly. An image pipeline processes scanned and photographic documents with three parallel models: a Tesseract engine for optical character recognition, a BLIP model for caption generation, and a CLIP ViT-B/16 encoder for semantic image embeddings, whose outputs are fused into a single textual representation. Retrieved evidence is passed to a large language model that generates the final grounded answer. On image- text retrieval benchmarks, the retrieval component reached Recall@5 of 92% and mean average precision of 0.89, with a mean query latency of 2.7 s. We describe the architecture, the evaluation protocol, and the observed trade-offs between accuracy and on-device cost, and we discuss the limitations that constrain mobile deployment.
Prashant Bhattarai, S. Baral, Kritika Thapa et al.· Journal of Engineering Issue...· 0 citations
Artificial intelligence has improved greatly and its demand in the education field has increased. To satisfy this demand, this research paper proposes an intelligent tutoring system which uses artificial intelligence techniques like Retrieval-Augmented Generation and Large Language Models. It aims to provide context-relevant and personalized help to students in the learning process. The system has a 3-layer architecture. The frontend layer is managed by React for smooth user interaction. The backend layer is handled by FastAPI for proper processing and response generation and the database layer is managed by ChromaDB, a vector database, which handles proper document storage and data retrieval. Students can upload their study materials such as textbooks, notes, etc. in various formats like PDF, DOCX, TXT. Then, the system processes the documents by recursively splitting text and generating embeddings to convert unstructured content into a proper structured knowledge base. After this, Retrieval-Augmented Generation helps retrieve or fetch the most relevant embeddings and combine them with user’s queries to generate contextually correct answers, based on user upload documents. This helps reduce hallucinations, which is the main aim. The system can also automatically generate flashcards and quizzes with adjustable difficulty. To evaluate the system’s performance, different document-based queries were tried and the system had 90% accuracy on average with 20%-30% decrease in hallucination (as compared to other systems). The response time is between 2 to 5 seconds. These results show the benefits of combining Retrieval-Augmented Generation with LLMs.
Azlaan Khan, Sakshi Chandekar, Atharva Baikar et al.· 2026 4th International Confe...· 0 citations