Predicting sudden cardiac death (SCD) in Chagas cardiomyopathy (CC) remains a challenge. An estimated 6 to 7 million people worldwide are affected by this condition. To address this challenge, we propose a novel intelligent system that utilizes multimodal and hybrid machine learning (ML) models to predict SCD risk in CC patients. We evaluated different approaches to combine clinical features (tabular data) and temporal features extracted from ECG-Holter exams. Our results indicate that the hybrid model (RNN with MLP) achieved the highest recall (sensitivity) of 91.63% and 81.08% of AUC. This study aims to develop a predictive model to support clinical decision-making in the context of SCD in CC. The model can potentially aid in risk stratification, enabling early intervention and potentially saving lives.