Bus transport if provided effectively can reduce private car usage and fuel consumption, alleviate traffic congestion, and save energy consumption. However, travelers care about waiting times, which depend on accurate information about bus arrival times. This information is influenced by various factors such as time, day, and distance. In this paper, we explore the design and implementation of a predictive model to estimate public transport demand in Nairobi, focusing on factors such as day of the week, time, and distance. We outline a comprehensive workflow that encompasses data preparation, feature engineering, data scaling, and model development. Due to the small size of the available data, we employed Conditional Tabular Generative Adversarial Networks (CTGANs) to enhance the quality of the dataset for better predictions. We compare the distribution of ground truth and synthetic data and use Long Short-Term Memory (LSTM) networks to predict bus arrival times. Our results show that the model’s performance improves when using a combination of ground truth and synthetic data. We also compare the performance of LSTM with Artificial Neural Networks (ANN) and other machine learning models, demonstrating the superior accuracy of LSTM in this context.