Voice AI

Annie Turner, Editor of Mobile Europe, recently interviewed Martin Rueckert, Chief AI Officer at Tallence, about Voice AI’s potential. Tallence operates important back-end services for telco companies in Europe, including operating some of the most important infrastructure for one of the largest German telcos, as well as being present in the Nordics and other European countries. Rueckert is the company’s AI lead, backed by a team of data scientists that embeds AI in telco infrastructure.

Rueckert began by explaining where he sees the biggest real-world impact for operators now. He said, “Voice will be the most important modality…for years to come…coming from customer service but now really taking it much, much further – so AI everywhere, voice everywhere in the telco and in the network.”

Turner noted that we are used to automation and chatbots. When asked what Tallence’s THOR Voice AI platform addresses that other approaches miss, Rueckert acknowledges that many telcos have already built voice agents for their B2B customers. However, this is complicated because “for Voice AI to work, you have to juggle all sorts of different technologies – like TTS [text to speech], STT [speech to text], voice detection – and also complicated large language models where you have to prompt them and integrate all kinds of back-end capabilities,” for which he gives examples.”

Tallence’s value to customers is to abstract that for the telcos, so that they can “build those products for respective customers without having to worry [about] all these individual bits and pieces that then construct Voice AI on the call.”

Tallence’s THOR platform is designed as a complete solution and Rueckert explains how it deals with the different elements, from authentication to monetisation. For example, operators handle sensitive information about and interactions with customers. Yet they must balance their use of Voice AI with trust and compliance regarding data protection and maintain transparency.

As Rueckert explains, this touches on many operational and technical aspects but they can be broken down into two broad categories. The first is voice AI for the telco itself. The second is the benefits it brings to customers who know telecoms is “a very regulated industry with clear guidelines…and…that is really where I would say they have a big advantage over Big Tech, which doesn’t have all those burdens.

Looking ahead two or three years, how will Voice AI change the way telecom companies interact with customers and their internal operations? “I expect telcos to actually make money with voice again; to sell services related to voice AI,” he says.

Then further down the line Rueckert expects a number of things to happen in parallel. These include telcos running voice AI-based services on behalf of their enterprise customers, as well growing demand for adjacent video-based capabilities.

He predicts such services “will massively take off, because they’re much closer to the [end] customer, much closer in terms of latency as well.”