International Research Journal of Education and Technology

The project examines the application of large language models (LLM) in automatic and enhancing data entry tasks. The system removes important information from unnecessary text, maps the data into a structured field, and detects errors in real time. The approach improves accuracy, scalability and flexibility in various industries, reduces human error and increases productivity. LLM for data entry project uses devices such as Pytessract for Parsing such as optical character recognition (OCR) and poppler for PDF parsing. The text extraction module removes raw text from various document formats, while the data mapping module converted the text into structured data, aligning it with predetermined areas. Document classification module classification classifies documents using the learning algorithm. The natural language interface module enables interactive input, improves access and efficiency. The project can reduce human error, increase productivity, and improve the overall efficiency of data admission processes.

Key Words: Large Language Model, Optical Character Recognition (OCR), Text Extraction, Data Mapping, Natural Language Interface.