Speech Technology Magazine

 

Splice Wins Over Customers with Human Voice Technology

Phrases, not words, make better customer connections.
By Michele Masterson - Posted May 14, 2015
Page1 of 1
Bookmark and Share

Splice Software has inked an agreement with reseller Clear Point Claims to provide U.S. insurance firms with personalized outbound voice messaging. While a reseller tie-up is not unusual, Splice's voice technology is.

The Calgary-based company uses data-driven dialogs to provide messaging services via inbound, outbound, and SMS as well as through email and the Web. Driving the technology is Splice Dialog Controller, which uses CRM data to create individual voice segments that are personalized for every end customer. The system is built with an API connected to other core systems that contain end-customer preferences and profile information, and it houses brand-related rules and permissions to meet compliance regulations in every state, province and country, explains Tara Kelly, CEO of Splice.

 "We called it [Splice Dialog Controller] because we believe that data should always be controlled and compliant," she says. "This is not a guessing game; there are hard-and-fast rules."

The insurance industry, a key vertical for the company, uses the technology to authenticate information, increase sales, and provide information that connects customers to claims. For example, when a customer is working with a broker, there might be an opportunity for the broker to obtain the customer's consent and permission, which can later be used to verify information, send important messages for catastrophic events, or cross-sell products. The system also takes in real-time data changes and updated policies to administrative systems.

Kelly says that what distinguishes Splice from other speech solution providers is a component of Splice Dialog called Splice Merge, which splices data together in phrases, not words.

"That's where we're very different than a Nuance or any other speech technology company," Kelly says. "We actually focus on recording from the beginning so all of our recordings are done at phrase libraries, not word libraries."

The system's dialog discovery determines the number of data elements that should be present in each phrase and then splices them together, similar to how phrases are built in Voice XML. The difference with Splice Merge is that it can accept changing data inputs on a real-time basis and match the phrasing. Splice has built out phrase libraries and holds the generic phrase for when date is not clean, accurate, or available.

 "We do not ever have just words recorded," Kelly says. "We always record in phrases so our messaging always sounds better. We've done hundreds of millions of calls across North America with customer interactions. We're also able to gauge the listenership by phrase, so we're able to map that as an engagement score and provide that back to the core system."

The company crowd-sources voice talent from all over the world in any language, dialect, or tone to provide customers with the most natural local speech, Kelly says. And it further personalizes outbound messaging by addressing customers by name.

Splice's unique system with real voices makes for a better listening experience, Kelly says.

"There's a million companies out there today that can send automated messages to update you with an API, but we're driving a better customer experience," Kelly says. "Customers listen longer to our messages because of the relevancy of the way we use the data and record a phrase—it sounds natural." 

Page1 of 1