The research results indicate that the UiPath Document Understanding system is capable of integrating AI, OCR, and Robotic Process Automation (RPA) technologies into a single automated workflow that is able to improve time efficiency, data accuracy, and reduce human errors in the banking document management process.
Intelligent Document Processing (IDP) using Artificial Intelligence (AI) enables organizations to efficiently process and analyze large volumes of unstructured and semi-structured data. Traditional rule-based and manual methods struggle with scalability and complexity, whereas AI-driven IDP leverages machine learning, natural language processing, computer vision, and deep learning to enhance accuracy and decision-making. This paper presents a comprehensive study of AI-based IDP systems before 2019, focusing on their architecture, methodologies, and applications in digital enterprises. It explains how technologies like OCR convert scanned documents into machine-readable formats and how supervised and unsupervised learning improve classification and data extraction. A modular architecture including ingestion, preprocessing, classification, extraction, validation, and storage is discussed. The study highlights improved performance in terms of accuracy, precision, recall, and processing time compared to traditional systems. Applications across banking, healthcare, insurance, and logistics are examined. Finally, key challenges such as data privacy, model interpretability, and system integration are identified, along with future research directions, emphasizing the role of AI-driven automation in digital transformation.
Andrew Collins, Emma Roberts· International Journal of Art...· 0 citations
Relevance. In the context of digital transformation and the active introduction of electronic document management systems, the volume of processed data is significantly increasing, a significant part of which contains personal information. An increase in the number of incidents related to data leaks, as well as stricter legal requirements in the field of their protection, necessitate the development of specialized tools for automated search and control of personal data in documents.
Goal. The purpose of the work is to automate the process of searching for personal data in the document management system based on the analysis and selection of effective methods for processing text information.
Research methods. The work used methods of analysis of unstructured data, template search based on regular expressions, as well as elements of text mining.
Results. In the course of the research, the existing methods of searching for personal data are analyzed, the choice of approach is justified, a prototype of the software module is developed and the architecture of its implementation is proposed.
S. Gavrilov, A. V. Shamsutdinova, I. F. Shaymardanov· Informatization and communic...· 0 citations
In this work, presented is a technical description of a software module for automating customer order processing in an industrial company. Analyzed is the finding that manual processing of unstructured orders takes twenty to forty minutes per document and leads to errors when matching against a catalog of four thousand items. Investigated is the microservice architecture based on FastAPI, comprising a RAG service, telemetry collector, and an embedded widget for the Bitrix24 CRM system. Revealed is that hybrid search combining semantic vector representations and a lexical inverted index provides high accuracy in matching customer descriptions against catalog nomenclature. Studied is the application of a three-level text extraction strategy for documents in PDF, Excel, CSV, and ZIP formats with Tesseract OCR fallback. Determined is that text chunking with overlap and a multi-level LLM API invocation strategy enable reliable extraction of product items from orders of arbitrary volume. Established is that fine-tuning the E5-base model on a domain-specific corpus using triplet loss with hard negatives improves semantic search quality. Formed is an iterative accuracy improvement mechanism through a closed-loop telemetry collection and manager feedback system. Proposed is a two-stage nomenclature matching scheme with gate optimization and caching to accelerate processing of typical queries. Developed is integration with Bitrix24 and 1C systems automating the complete cycle from file upload to order creation in the accounting system. Substantiated is the advantage of the proposed approach over manual processing and classical full-text search in both speed and result quality. Presented is a description of validation results on real customer data confirming a three-to-four-fold reduction in order processing time.
A. A. Erofeev, A. N. Babkevich, M. Ozerova et al.· International Conference on...· 0 citations
This study focuses on creating and implementing a personal finance management information system that incorporates voice recognition technology integrated with Google Gemini AI 2.5 to improve the efficiency and precision of recording financial transactions. Traditional systems encounter challenges such as labor-intensive manual inputs, elevated error rates in records, and inconsistent user engagement. The Waterfall model was utilized, covering requirements analysis, system design, implementation, testing, and ongoing maintenance. Data gathering involved conducting interviews with 10 participants, observing recording methods, and reviewing literature related to voice recognition and AI technologies. The application was created as a hybrid mobile app utilizing Apache Cordova and the Web Speech API for converting speech to text, alongside Google Gemini AI 2.5 for the automatic extraction of transaction details (type, amount, category). Data is stored locally using SQLite to enable offline access. Blackbox testing, accuracy assessments, and user acceptance evaluations resulted in 87% accuracy in voice recognition, 92% accuracy in transaction categorization, an 83% reduction in input time, and a System Usability Scale (SUS) score of 78, reflecting strong usability. The SRIPSI system allows for transaction recording through everyday Indonesian/English language, a real-time balance dashboard, transaction history, and straightforward reporting, effectively addressing the shortcomings of previous manual applications like Money Lover and Wallet.
Siti Nurul Hidayati Aurel P. W, Tholib Hariono· NEWTON: Networking and Infor...· 0 citations
The increasing demand for efficient administrative services in higher education institutions has highlighted the need for effective correspondence management systems. Conventional mail handling processes often result in delays, misrouted documents, difficulties in archiving, and challenges in information retrieval. This study aims to develop a web-based Smart Mail Management System by integrating Optical Character Recognition (OCR), rule-based intelligent routing, automated notifications, and digital approval features to improve correspondence management within the Faculty of Science and Technology, Universitas Ibrahimy. The research employed a Research and Development (R&D) approach using the Waterfall Software Development Life Cycle (SDLC), which consists of requirements analysis, system design, implementation, testing, and maintenance. Data were collected through observation, interviews, documentation analysis, and literature review. OCR technology was utilized to convert scanned letters into digital text, while a rule-based mechanism automatically classified and routed correspondence to the appropriate recipient. The results indicate that the proposed system reduces manual processing, improves routing accuracy, accelerates document handling, enhances document traceability, and facilitates centralized digital archiving. Furthermore, the integration of automated notifications and digital approval features supports a more efficient administrative workflow. These findings demonstrate that the developed system contributes to improving the effectiveness, accuracy, and efficiency of correspondence management while supporting digital transformation initiatives in higher education institutions.
Wardhatun Khasanah, Abd. Ghofur, F. Santoso· G-Tech· 0 citations
Traditional document management systems suffer from inefficiencies in organization, retrieval, and data extraction, often relying on manual entry and rigid field structures that fail to accommodate diverse document types. This project proposes an AI Smart Document Management System with Dynamic Field Extraction that leverages Deep Learning and modern web technologies to address these limitations. The system supports a broad range of document categories including invoices, contracts, medical records, and academic transcripts enabling intelligent, automated extraction of key fields without requiring predefined templates. Optical Character Recognition (OCR) powered by
Tesseract processes uploaded documents, while a Large Language Model (LLM) running through the Groq API performs context-aware, dynamic field extraction and document summarization. All documents are indexed and stored in MongoDB, with vector embeddings enabling semantic search across the repository. Users can perform natural language queries to retrieve relevant documents efficiently, bypassing traditional keyword-based search limitations. The backend is developed using FastAPI for scalable and asynchronous API handling, while the frontend is built with React.js for a modern, responsive user experience. Authentication and session management are secured using JWT tokens. The system further incorporates AI-assisted summarization and an intelligent assistant for document-level querying. By combining OCR, LLM-driven extraction, and semantic retrieval, this solution significantly reduces manual processing overhead and modernizes organizational document workflows.
Keywords:
Smart Document Management, Dynamic Field Extraction, Optical Character Recognition, Large Language Models, Semantic Search, Vector Embeddings
Kata Raju Reddy, Korada Ramya· International Journal of Sci...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
