We'd like an AI driven function to derive learned values for activity durations at activity type and other attribute levels within work order. This should learn at at an area and individual engineer level. The data needs to be relevant and good to use, so should consider volume of activities undertaken, as well as exclude outliers.

This should then be applied automatically (daily but weekly would do) to existing and newly booked appts. Should be applied at area and tech level, the area values aggregated from techs within the area.

This should be a continuous cycle.

These values should be used within the RSO function to understand where best an appointment is placed based on things like reducing travel, maximising utilisation etc.