Inbound Communications Intent Prediction for the Largest Telecommunications Provider

Image
Case Study: AI/ML Solutions

Inbound Communications Intent Prediction for the Largest Telecommunications Provider

SUMMARY

One of the largest U.S. telecommunications provider wanted to realize significant cost savings by enabling AL/ML at call centers to prevent calls from reaching call center agents. Xtrictech 3PL created a solution to enable customers to self-serve via its IVR by playing verbal messages. The solution successfully deflected 3.8 million calls during the first year of deployment.

Challenges

  • Client receives over 6 million calls at its call centers monthly. The caller first interacts with an Interactive Voice Response (IVR) system where most rings have to be transferred to call center agents.
  • To realize significant cost savings, the client wants to enable self-serving via its IVR by playing verbal messages to prevent calls from reaching call center agents.

Solutions Offered

  • The AI/ML solution comprises a group of real-time prediction models to determine why the customer is calling. The solution generates probabilities on over 280 intents and arbitrates the most likely intent of the caller in real-time. The dreams predictions are used to drive voice messages played to the customer while at IVR (ex: are you calling about this “charge” on your last month bill?), offering a self-serving option(s) to resolve the customer need/inquiry, thus deflecting the call of being transferred to a call agent.
  • The solution successfully deflected 3.8 million calls during its first year of deployment.