Has anyone setup multiple task intelligence models for field prediction and auto fill in case one models doesn't behave well enough with a broad training and prediction filter? Are you using TI in combination with static assignment rules for the most obvious keywords etc?
Any special tweaks you were forced to apply to get better results?
We are trying to de-customize by removing our own assignment prediction models to move over to TI.
So far only 135 post in the community mentioning Task Intelligence so it's perhaps not widely adopted (yet?)
