If it was possible to collapse multiple trees into a single decision tree, without loss of performance, the same effect should be achievable with a single decision tree. The trees constructed by boosting algorithm take different paths, i.e. choose different variables to make splits, make different splits, etc, there's no way to losslessly combine them into a single tree.
Algorithms such as random forest or boosting are not very computationally demanding at prediction time. If they are for you maybe you should change the hyperparameters, for example, use trees of smaller depth, print them, use less trees/iterations. If this still doesn't help because you have some tight constraints, consider using a simpler algorithms like logistic regression, single decision tree, or naive Bayes classifier.