yeah this is pretty trivial to do .
basically you create a voice with N channels, with a midi (note) out then use voice index to determine which channel to go to.
(alternative you can use a keyb zone if you want to do more traditional splits)
only thing which is 'slightly tricky' is... the voices are independent so if you want to base your logic on what the other voices are doing, then you need to use a table in the parent patch to hold this data... not hard, but if your new to Axoloti its perhaps not obvious.... but i think for what you describe this wont be necessary.
note: this is the simplest way, and has a limitation of dealing with up to only 16 notes
you can also do this by using the midi script object, the complexity of your script then is then down to what you need to implement... and its not visual patching, so some will find this tricky.