International Journal For Multidisciplinary Research

E-ISSN: 2582-2160     Impact Factor: 9.24

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Call for Paper Volume 8, Issue 3 (May-June 2026) Submit your research before last 3 days of June to publish your research paper in the issue of May-June.

SmartScan - Multi Agent AI System for ID data extraction in combination with OCR engine.

Author(s) Ms. Meera Sawalkar, Mr. Prathamesh Gokulkar, Ms. Angel Londhe, Ms. Isha Badhe
Country India
Abstract Artificial Intelligence (AI) and Natural Language Processing (NLP) have transformed how enterprises handle vast amounts of unstructured financial documents such as invoices, receipts, and reports. Conventional OCR-based extraction systems, though functional, struggle with document diversity, semantic ambiguity, and the absence of reasoning. Recent advances in multimodal deep learning models such as LayoutLMv3, Donut, and LongFin have enabled deeper understanding of document structure and meaning. However, gaps remain in combining extraction accuracy with interpretability and user interaction. This paper presents a comprehensive survey of AI techniques applied to financial document intelligence and introduces our unique contribution: a full-stack, multi-agent financial assistant capable of intelligent extraction, validation, and conversational reasoning. The study consolidates findings from major research papers, explores theoretical foundations, evaluates datasets and models, and identifies challenges and opportunities for the future of explainable and interactive financial AI.
Index Terms—Financial AI, Document Understanding, Deep Learning, LayoutLM, Information Extraction, Donut, Conversational AI, Multi-Agent Systems.
Field Computer > Data / Information
Published In Volume 7, Issue 6, November-December 2025
Published On 2025-11-15
DOI https://doi.org/10.36948/ijfmr.2025.v07i06.60643

Share this