This project aims to systematically characterize intangible cultural heritage including mak yong,  wayang kulit, bangsawan and silat as documented in Malay-language periodicals from 1930 to 1960, a period spanning colonial rule through early independence in Malaya / Singapore. At its computational core, the research seeks to develop AI models for Jawi OCR transliteration, fine-tuned BERT classifiers for cultural content categorization, Vision-Language Models for heritage image recognition, and Retrieval-Augmented Generation systems for large-scale extraction from digitized archives. The goal is to produce a suite of open-weights models and structured datasets that will make previously inaccessible Jawi-script sources searchable and quantitatively analyzable at scale, establishing a replicable framework for multilingual heritage research across Southeast Asia. More information about the project is available at https://culturalheritagenus.github.io/jawi/.