loading

Business Impact

  • 78%

    Queries handled by virtual agent

  • 88%

    Accuracy in user authentication

  • 40%

    Reduction in call duration

Customer Key Facts

  • Location : North America
  • Industry : Government

Problem Context

The customer manages over 6,100 public housing units and is the fourth largest administrator of all housing programs in the state of Texas. They receive 150,000 calls each year in English as well as Spanish with regards to housing applications. These application status queries, which form 40 percent of the total call volume, require the housing authority representatives to manually authenticate the applicants by asking a series of questions. This time-consuming task led to longer wait times and frustrated callers.

Challenges

 

  • Performing user authentication on call for application status queries
  • Reducing the number of calls that will be forwarded to Live Agent
  • Handling nuances of telephonic conversation in English as well as Spanish
  • Connecting Google Dialogflow with the client’s telephony provider
Challenges

Technologies Used

Google Cloud Platform
Dialogflow
GitLab
Node.js
Jive
Voximplant

Resolving Queries for a Fraction of the Cost of Hiring Additional Staff to Handle Inbound Calls

The representatives had to be trained and skilled to handle queries in both languages, which led to increased effort and cost. Due to restricted contact hours, representatives were unable to respond to user queries outside business hours and every time a user was connected to a new representative, they had to repeat the details of their previous conversations. Therefore, the customer wanted to better address customer queries and reduce the workload of its representatives by transforming their existing telephony IVR service.

Solution

Quantiphi built a multilingual, telephony virtual agent solution with smart agent routing and conversational history API; assisting representatives in handling high call volumes and enabling process optimization and cost effectiveness. The virtual agent authenticates users, with the capability of addressing 80 percent of all housing application status queries in English and Spanish, and connects them to the specific representative directly based on previous conversations.

Result

  • Reduction of the Average Handling Time from 5 minutes to 3 minutes
  • 24/7 user support even during non-business hours and public holidays
  • Conversation designed for recognizing nuanced pronunciations of caller names while maintaining seamless user experience
  • Smart agent routing allows a user to connect to a specific representative who has worked on their case before
  • Conversation history API designed to facilitate storing important user parameters during call for better followups on user queries

Looking for similar project?

Let's Talk

Get your digital transformation started

Let's Talk