I'm not an engineer so I have no idea if this is a difficult thing to develop or not. My guess is that it's not that difficult as it is not sound based but more of an algorithm thing.
So what about a waveform generator, that could be applied to different parameters and effects of the kemper (volume, frequency, filter and yes, tremolo...). It would allow user to choose between different waveforms of course, sine, wave, saw, square, triangle, random, etc. and also it would be tempo synced with the choice of different subdivisions.