Level 7 Systems wins customers with Watson Speech to Text API

By | 2 minute read | January 31, 2017

One of the world’s most disruptive industries is voice services. Once dominated by a few large and established players, new innovations in Internet Protocol (IP) communication technologies are making it easier for non-telecom businesses—including software and hardware companies—to compete effectively. To be successful, however, it’s not enough for these challengers to have a great idea. They must also be able to focus their efforts on what they do best and get their solutions to market quickly.

Level 7 Systems is one such startup. Founded in 2008, the London, England-based Voice over Internet Protocol (VoIP) provider develops enterprise and residential communication solutions, such as their flagship service, VoIPstudio. As a relative newcomer in the highly competitive global telecom market, Level 7 must respond rapidly and efficiently to customer needs—or lose out to rivals who can.

Identifying untapped business potential

Level 7 Systems was seeing an uptick in the number of customers requesting text transcripts of their voicemail messages, a promising opportunity for a new service offering. The company first attempted to build its own speech-to-text platform, but before long, this resource-intensive task grew complex and out of scope.

“We needed to concentrate on our core business,” says Chris Maciejewski, Level 7 Systems’ co-founder and technical director, “and not on getting our resources involved in intricate technical issues.”

Taking advantage of opportunities with Watson

The company researched several third-party solutions before choosing Watson Speech to Text, an easy-to-use service that applies IBM’s speech recognition capabilities to convert the spoken word, in multiple languages, into text. Watson’s cognitive technology applies neural network processing, combined with a deep understanding of grammar and natural language, to translate unstructured data like voicemail messages with far greater accuracy than non-cognitive speech-translation products.

“We chose Watson because of the quality of the actual voice-to-text conversion, and how well it integrates with our systems,” Maciejewski says.

By leveraging the self-learning capabilities of Watson’s Speech to Text application programming interface (API), Level 7 Systems was able to offer a key feature that allows customers to report transcript errors. This enables the API to improve its translation accuracy rate over time. And because the Watson API is so easy to use, Level 7 could deploy its solution in just seven days.

Boosting the bottom line

Level 7 Systems attributes six to seven percent of its business growth to the straightforward and powerful Watson Speech to Text API, one of many cognitive IBM Watson APIs and helpful developer tools available on the Watson Developer Cloud.

To learn more about how Level 7 Systems employed Watson to grow its business, win new customers and create a competitive advantage, read the case study.