The evaluation of student answer scripts holds immense significance in education systems worldwide. However, the traditional evaluation process can be time-consuming and demanding for teachers. Consequently, teachers face difficulties in providing detailed explanations to each student regarding their marks on every question and answer. Moreover, students often lack constructive feedback, which hinders their ability to improve and identify areas of strength and weakness in their work. To bring more transparency to the evaluation process, it is essential to provide students with review comments for each question, rather than just giving them a final score for the exam. This approach allows students to have a clear understanding of their performance on individual questions and enhances a better learning experience. To address these challenges, this research explores the utilization of advanced techniques such as Optical Character Recognition (OCR) to digitalize handwritten text from assignments, internal, or final exam answer scripts. Large Language Models (LLM) like GPT-3.5 and LangChain framework are then used to evaluate the digitalized student answers by comparing them with predefined evaluation criteria. To ensure that the automated evaluation system is reliable, multiple patterns of questions and answers were tested using data from real students answers in practice tests from higher secondary school subjects such as Social Science, Mathematics, and Science subjects. This automated evaluation system aims to streamline the evaluation process, providing comprehensive feedback to students while reducing the burden on teachers. Furthermore, it significantly reduces evaluation time by 90%, providing an efficient and cost-effective solution.