One can build any custom model in Sparkflows as long as the library being used by Sparkflows is installed in the python environment. If a library which one wants to use is missing, it can easily be added to the python environment by activating the environment and then running a pip install on the package.