Automating Template-based PDF Customization using Large Language Models

Automating template-based PDF customization is essential across industries such as finance, healthcare, and legal documentation. Traditional scripting-based methods often lack flexibility and scalability. This research investigates the integration of Large Language Models (LLMs) to enhance the generation and modification of structured PDFs. By leveraging techniques such as prompt engineering, Named Entity Recognition (NER), and layout-aware processing, LLMs enable dynamic content mapping onto predefined templates while maintaining coherence and accuracy. Comparative analysis against rule-based approaches highlights LLMs' advantages in adaptability, processing efficiency, and reduced manual intervention. Experimental evaluation demonstrates improvements in content accuracy, formatting precision, and processing latency, validating the potential of LLMs for scalable and intelligent document automation.