DialogTech Launches Conversation Insight 2.0, Powered by Nuance Transcription Engine

Call attribution technology provider DialogTech has launched the second iteration of its audio parsing and dialog analytics solution, Conversation Insight. Conversation Insight 2.0 now incorporates Nuance Communications' Transcription Engine, resulting in increased speed and an accuracy boost of more than 50 percent compared to DialogTech's previous iteration.

DialogTech was known as Ifbyphone until last year following its acquisition of Mongoose Metrics. Its stated mission is to bring sophisticated call and conversation analytics to distributed sales teams. "These kinds of functionalities have never been available before outside of the call center," says DialogTech CEO Irv Shapiro.

Joining the trend of marketing service solutions for smaller businesses, DialogTech seeks to provide powerful call attribution capabilities to what it sees as an underserved market: smaller businesses, franchisees, and remote agents, none of which enjoy the benefits of the centralized call center interaction data associated with larger enterprises.

"The general wisdom these days is that you should never talk to your customers, that mobile and self-serve are going to dominate the market. But think of auto dealers and insurance sales forces," Shapiro says. "Their first customer touch point is often a phone call, not a smartphone app interaction or web engagement, and that call more often than not isn't routed through a call center. These businesses haven't traditionally had access to the enterprise-grade call attribution solutions that we're providing. Before us, there was a significant segment of actionable marketing and customer service data that was being lost."

Now with the Nuance Transcription Engine, Conversation Insight 2.0 can map and provide data on customer journeys at an exceptionally granular and searchable level. Leveraging the Nuance Transcription Engine and technology by Elastic (formerly Elastisearch), Conversation Insight 2.0 allows users to search their entire database of call audio for keywords and create sophisticated templates on transcription-based parameters, all while providing audio tracking back to any original media source, deep data on mobile, geolocation, and a cross-reference to the user's call system of record.

The new version of Conversation Insight also provides a new interface that separates speakers into discrete waveforms, with the ability to auto-mark portions of audio that have been identified with user-specified keywords for those looking to drill deeper into conversation context.

Conversation Insight 2.0 is also running the Nuance Transcription Engine in Amazon Web Services using an auto-scaling group from zero to 90 eight-core servers, ensuring speed while maintaining maximum accuracy. "We only use the high-compute model," says Shapiro, noting that Conversation Insight is configured to accurately transcribe six hours of audio in minutes.

The new search template functionality will be available via the DialogTech portal immediately, and through API within the next few months.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues