Service Request Triage – Demo

Video demonstrates the RPA Category of Service Request Triage.

Service Request Triage is a type of RPA for automating office processes that generally involve firstly, the process of classifying what the request is, or diagnosing the issue, and secondly actioning that request with a next action. To demonstrate this, the Video demonstrates an RPA robot processing Customer Email for a fictional furniture manufacturer.

It shows the RPA robot categorising Customer Email and conducting the next action for a:

  • Late delivery update
  • Request for a Scheduled delivery date
  • An issue with Damaged goods
  • A Customer complaint
  • Requesting a return
  • Stock availability question

This demo is divided into 2 parts. The first part demonstrates the actual RPA robot conducting the Customer email automation, and second part, gives more detail behind the automation, and explains how RPA robot uses AI and specifically, Machine Learning, in order to recognise different categories of customer email.

Key Points:

  • In a hospital’s accident and emergency department, a triage nurse would first diagnose the patient’s health issue and then determine the next action. Likewise in a business setting, a human employee is first diagnosing what the customer’s request is and then actioning that request.
  • Generally an IT service request for example, generally involves the IT technician diagnosing the user’s issue, and then determining the next course of action to rectify it.
  • Similarly, the process of Returns or Refunds might first involve determining the type of return requested before actioning this for the customer.
  • And Customer Service Email (which is this demo), generally involves an agent reading the email to determine the customer’s request, and then following it with a next action.
  • 80/20 : What’s interesting about these type of office processes, is that typically, a few types of requests account for the large majority of the transaction volume, a classic 80/20 if you will. So automating the few high volume requests can make a dramatic impact on agent workload.