We couldn’t be happier to announce that PolyAI has been included in a second Gartner report this month. Emerging Technologies and Trends Impact Radar: Customer Analytics for Customer Experience (October 2020) takes a deep dive into customer analytics solutions and is well worth a read for any company hoping to exploit such technologies in the near future.
The report looks at many areas of technology supporting customer analytics, from IoT to customer journey analytics to smart personalisation and voice assistants.
Why is Voice so Difficult?
The report notes, “being able to offer multiturn voice is a very complex task for many reasons; classifying intent from voice is hard because speech-to-text provides an imperfect signal and people speak idiosyncratically, making mapping language to intents an untenable approach. Furthermore, carrying context becomes more complex with every conversational turn, making traditional tree structures an unsuitable architecture.”
PolyAI has developed the ConveRT model, in part to allow our voice assistants to classify intent, even when speech is incompletely recognised. Just as we humans are able to understand what someone is saying, even if we miss the odd syllable here or there, PolyAI voice assistants are able to use a vast amount of historical data to best predict what a user has said should speech-recognition fail.
Because the ConveRT model has been trained on billions of conversations, it’s able to understand idiosyncratic language, giving end-users the freedom to speak naturally.
Our proprietary content programming framework means we don’t rely on traditional tree structures, allowing conversations to flow freely without compromising short-term memory or context.
The Future of Voice
Due to the difficulties of developing voice assistants (as highlighted above), uptake of such technologies has been slow. But over the next few years, voice automation will become commonplace.
“As a horizontal application, voice assistants will reach across multiple industries and markets” states the Gartner report. “Industries that are mainly focusing on servicing the consumer markets, such as hospitality, travel, retail, finance, telecom (among others) will see the biggest adoption.”
There is no question that voice assistants will play a key role in the future of customer service. Here at PolyAI, we’re utilising over a decade of academic experience to bring state-of-the-art conversational technologies to commercial customer service applications.