Visually distinct rock art at Khao Chan Ngam in Thailand's Khorat Plateau presents a rare opportunity to investigate how rock art may have represented identity in the past. However, the process of analysing and then categorising rock art can be a time-consuming and labour-intensive task. Automating this procedure would allow analysis to be scaled, enabling this research to be applied in areas with little previous research, such as in Thai rock art. Recently, Machine Learning (ML) approaches have been used to recognise and extract distinctive characteristics from a variety of artistic and cultural media including rock art. In this presentation, I will describe research I am undertaking to provide the first ML-based study of Thai rock art using convolutional neural networks (CNNs) and transfer learning to classify images. These techniques will be used to perform supervised and unsupervised learning to identify attributes that may be used to categorise rock art in both small and large image collections. Utilising Machine Learning in rock art analysis may enable large-scale analyses with new approaches to feature selection, classification, and clustering of rock art styles, contributing to recognising identity, cultural association, and variation in rock art. Lower-code (software platforms that requires little or ‘low’ code to analyse data or drag-and-drop coding for non-specialists) options for Machine Learning are becoming increasingly available and this study will show the useability and adaptability of this approach for archaeological research. A further outcome of this case study will be an introductory guide for ‘getting started in Machine Learning’ for archaeological purposes.