FREE OCR IMPLEMENTATION WITH GOOGLE CLOUD VISION AND TELEGRAM FOR ACCOUNTING APPLICATIONS

Abstract

Digital transformation in accounting practices demands high efficiency in processing transaction data from physical documents to digital systems. This study explores the implementation of Optical Character Recognition (OCR) technology using the Google Cloud Vision API integrated with the Telegram Bot platform for free automation of accounting data input. The main problems faced by micro and small enterprises (MSMEs) are the high cost of automated accounting software and the complexity of manual input that is prone to human error. The methodology used is the development of a serverless- based system that connects the Telegram interface as a document scanner with Google's machine vision for text extraction. The results show that this integration is capable of extracting data from receipts and invoices with an accuracy level above 90% under normal lighting conditions. In conclusion, the use of free tools and cloud-based infrastructure can democratize advanced accounting technology for small economic actors, reduce administrative workloads, and increase the validity of financial reports.