In Sparkflows is it possible to build a ml model in one machine and use it for prediction in another machine

Yes. It is possible.

Sparkflows supports 'Spark ML Model Save' node using which one can save a model into a file.

Generated file can be copied over to another machine.

Sparkflows supports 'Spark ML Model Load' node using which one can load an existing model from a file. Which can be using for prediction using Prediction nodes.