I would look at yeast strain and fermentation temp scheme to address your diacetyl.
I agree with this. Pick a low D producer that flocs quickly and rouse often as you bump the temps. You should probably incorporate a forced diacetyl test before you cold crash as well to ensure that you won't run into issues down the line.
The rousing part I did not know. Does it really help? And the forced diacetyl test (you mean rest, I assume?), who would that work if fermentation should be done after two and a half days?
Well, both solutions would likely blow-up your target drinking time. Rousing can help if you have yeast flocculating well before the wort is attenuated to the desired degree. A forced diacetyl
test, you warm a sample more to room temp, while covered with a coaster or somesuch, swirl, smell, taste. You are "forcing" the diacetyl to become more discernible at warmer temperatures. Again, if confirmed you then need the rest, which moves the goalpost.