Design and Development of a Web-Based Tax Invoice Recapitulation Application Using the Optical Character Recognition (OCR) Method

The process of recapitulating tax invoice data is crucial because it directly relates to a company's financial management. This process can be challenging for business owners when implemented using inefficient methods. For example, at CV Banten Inspirasi Mandiri, tax invoice data recapitulation is still carried out conventionally by manually entering data into Microsoft Excel. This approach is ineffective in terms of time efficiency and offers limited guarantees of data accuracy. To address this issue, an Artificial Intelligence (AI)-based application was developed to optimize the company's time and resources. Specifically, this system utilizes Optical Character Recognition (OCR) technology, which enables automatic data extraction from tax invoice files. The implementation results showed a time efficiency rate of 92.92%, with 100% accuracy for OCR results from PDF files and 89% for OCR results from JPG files, outperforming conventional methods, which only achieved 80% accuracy.