Voice recognition is an established tech component in everything from virtual assistants to cars, but sometimes it can be a bit spotty—just ask anyone with a thick accent.
Fluent.ai is improving the voice recognition game—one garbled sentence at a time—as more multinational companies turn to the Montreal-based organization for its state-of-the-art platform. Fluent.ai created the world’s first acoustic voice interface for intelligent devices and user interfaces. It can be used with consumer-electronic OEMs and telecom service providers around the globe.
The product is called the Fluent AI Engine and uses a neural network and machine learning to learn languages on-the-fly, just like a child might. It associates specific sets of sounds with an intended command, which means Fluent AI is the first entirely acoustic voice interface. This gets rid of the complex and often inaccurate step of converting speech to text, which is what Amazon’s Alexa or Apple’s Siri does.
Fluent AI ran a few tests against other AI voice assistants to see how it measured up. In the first test, Fluent.ai used a small and specific vocabulary, working offline, against a bigger speech-recognition technology connected to the cloud. The second test had both systems working offline with the same vocabulary.
Normal speech-recognition platforms claim around a 95 per cent accuracy rate, but only in ideal environments, such as low background noise and a clear unaccented American English speaker. In the two tests that were run, Fluent AI used speakers with accents in a fairly noisy environment and maintained accuracy rates of 96 per cent. The other competitor? Only 72 per cent.
Acoustic voice interfaces work wonders when it comes to updating them with new languages. In a case where Fluent AI was tasked to create a multilingual automated-voice-response system, the platform took 10 weeks to create functioning custom models for multiple languages as a proof of concept. This would normally take other systems, several months, or even years.
“Fluent.ai demonstrated how customer-service operations can quickly roll out AVR support for additional languages at a fraction of the cost, and in a fraction of the time, compared with conventional voice-assist platforms, for a much shorter time to market,” said Niraj Bhargava, CEO of Fluent.ai. “For our telecom customers, this extends beyond call centre operations to include smart devices for the connected home and the Internet of Things.”
Perhaps the most impressive aspect of the Fluent AI Engine is that it can understand everyone. Thick Scottish brogue? No problem. Eloquent Southern twang? Fluent.ai has your back.
“Consumer needs often lie with a platform that can quickly be taught to understand specific language for a specific context,” reads a statement from the company. “And most importantly, do so with at least 96 per cent accuracy – every single time and regardless of language, accent, speech impairment or background noise. That’s Fluent.ai.”
Earlier this year Fluent.ai received $1.8 million in seed financing to accelerate commercialization of their platform. The company is working with several large organizations and the exact names and nature of their partnerships will be unveiled soon.