Maximizing Efficiency in Public Employee Retirement Services with Conversational AI and Voice Biometrics

Customer

The company was founded in 1920 with 13,331 members. Today, it has grown into the largest municipal public employee retirement system in the United States with more than 350,000 active members and retirees including civilian employees such as clerical workers, accountants and social workers, and uniformed employees such as New York City Correction Officers and Sanitation Workers.

Challenge

The company was looking for customer experience improvements for their call center IVR and to automate more of their IVR services:

Automate routine tasks like authentication, and Frequently Asked Questions

Automate SMS links to callers for commonly requested forms and information

Improve authentication security utilizing Voice Biometrics

The company previously had just one agent queue for all customer services requests, and no automation in the IVR at all. Agents were frequently spending time and effort manually authenticating callers directly, and also answering requests for forms and common questions, instead of spending their time resolving more challenging or unique caller concerns with their retirement accounts.

Solution

Previous to Omilia’s engagement with the company, the only answer to these problems was to hire more agents or allow caller wait times to increase.

Omilia presented an alternate solution: using a seamless Conversational Voice experience built on Omilia Cloud Platform, fulfill caller requests via automation using a custom-built intent model, including automating SMS responses to numerous caller requests, and authenticating calls directly in the IVR.

The custom-built Intent model that Omilia deployed is able to easily identify dozens of different kinds of caller requests. Customers get exact answers to specific questions in a matter of seconds. This also empowers the company to make better decisions on which types of requests should be serviced by the conversational AI and which should be given to agents due to their complexity, maximizing their agents’ time serving more important needs.
 

Result

Leveraging Omilia’s Natural Language IVR, the company has been able to authenticate over 77% of callers in the IVR, reducing handling time with agents and allowing callers to begin self-service experiences within the IVR, which will continue to drive containment increases as self-service functionality is deployed.
 

Data from Aug 1-31, 2023:

Caller Authentication Success Rate

Of the failures, 82% occur when asking for the Member Number, which some members may not remember and enter incorrect numbers or no number at all

Phone Number Gathering for SMS Deflection Success Rate

What’s next

With the Omilia Conversational Voice application actively learning from customers and serving customer requests, the Omilia delivery team is continuing to develop and deploy additional self-services to further increase IVR containment for the company.

Additionally, the company will be adding Voice Biometrics authentication to their Omilia IVR in conjunction with an integration to LexisNexis for validation of callers, ensuring a smoother and easier process of identifying callers and serving the purpose of their call quickly and securely.