Experience IT developed Call-AI - a call centre and resource scheduling solution which enabled the AI division to consolidate 10 branch offices into a single call centre. It also enabled the efficient management of the technician resources in a business which was highly seasonal and where such efficiencies made big effects on the bottom line performance.
When a customer calls to ‘place an order’ their details are automatically presented on screen even before the call is taken. The operator can record the details of the request in seconds (the only options for a service call are today/tomorrow am/pm - when Daisy is ready she doesn’t wait around!). The farmer may optionally indicate the number of cows to be inseminated and the preferred bull/breed. Calls are automatically assigned to a technician assigned to the ‘cell’ or area on the requested day as the entire system is geo-aware. The person in the call centre does not need to know to whom the call would be assigned (though they can see the details of the auto-assignment). A notification, including a service call ID is sent by SMS to the designated technician.
The final solution was a true ‘expert system’. It took over from people who knew the customers personally (and therefore where they lived) and the technicians operating in their area (and thereby which technician to assign the call to). Call-AI was good for the technicians too as previously notification of call-outs was via VHF 2-way radio. Technicians are ‘indisposed’ for lengthy periods due to the nature of their work. It is much simpler to review text messages on a mobile phone on return to the car. Listed below are some of the technology enablers used to deliver this project:
- TAPI to link the PC software to the Ericsson MD110 PABX (for CallerID)
- SMPP to enable the SMS messages to be output via an SMS service directly to the mobile operator SMSC. Regular MO (mobile originating) SMS (even using a dedicated SMS modem) was not sufficiently fast (or resilient).
- The rostering module was based on a multi-level ‘map’ which had many cells each representing the area which was serviced by a single technician. The geo-coded cells grew and shrank according to the time of year from a peak of 70 cells (= 70 technicians on duty) to approximately 15 in the quite times of the year.
- Call-AI was linked to the client’s VAX based billing platform to reconcile between scheduled calls and invoices and to synchronise stock information.
