It seems to me you've got a few options which you can do incrementally and prove each step.
First I'd consider just using a DIN unit and a core, and create a button for the first 3 fingers on each hand. This would enable you to play anything in C and get a feel for how it would work at minimum cost. Even after I would not model the bottom B/Bb keys, and for a simple prototype I'd also probably ignore the bottom C key, and left hand high D/D#/E/F/F# keys. Instead of a single octave key I'd recommend you have some down octave-up octave keys where you would normally put your thumb. This is the way the Yamaha WX5 works, and it can give you something like 8 octaves. Your best bet is to then work out in software which note to output depending on the keys and the octave keys which would be a lot easier than doing it in hardware. Just as a starter you could mount the buttons on a broom handle or similar just so you can check the software. This would give you a useful and cheap prototype. The WX5 also has a thumbwheel where your right thumb would normally sit on the sax, and you could assign this to pitch wheel or modulation wheel. Once you are happy with all of this at minimum cost, you could then think about having it on a more ergonomic body, adding the left hand high key cluster, and how to deal with the breath note-on/note off. I suspect breath velocity, and trying to simulate the lip bend will be more difficult. I'd probably leave the lip bend until last. Also be warned that you get a lot of queues on how to lip your note through vibrations through your head, and you don't get this sort of feedback on a midi-sax, making tuning hard.
You could make the breath a simple on/off at the beginning and use one of the remaining inputs on the DIN module, just to see how it goes. Once you've got all that working then I'd think again about the air velocity and how to measure it.
If you're thinking about midifying a real sax, then you could use the same software for deciding on the note played but with actuators on the main keys. But the challenge here would be trying to stop the reed buzzing.
I hope this gives you a more simple starting point.