Fit logistic regression to your dataset, then fit random forest to your dataset, and when you use them for inference with an independent dataset, average their predictions into one prediction. This is called ensemble averaging, and it is one of the simplest ways to benefit from fitting two different types of prediction models.