The aim of this project is to structure the digitalized samples and to automatically extract the label id, year, marker information and slide number of the microscopy slides. We have used Qwen LLM and finetuned based on manually annotated examples together with prompt engineering.