Amazon Connect

Amazon Introduces Customer Voice Recognition Feature to Call Center Platform

Amazon announced plans to integrating voice recognition technology into its enterprise call center platform at its yearly re:Invent conference. The call center service provided by Amazon Web Services will soon use artificial intelligence to analyze an employee or customer’s voice and quietly confirm their identity. The goal is to improve security without all of the extra steps usually required to determine who someone is on the phone.

Voice ID

AWS provides the call center platform to clients as a tool for interacting with customers while collecting and analyzing data on those conversations. Usually, businesses rely on asking identifying questions like birthdays, social security numbers, or previous addresses to check if a caller is who they claim. Voice ID is designed to skip the often tedious process and make it harder to steal an identity on a call. When an agent or a customer agrees to use Voice ID, the software uses a few seconds of the call to break down their speech biometrics and label elements like pitch, tone, and rhythm. Those bits are then saved and marked as an individual’s voice file for future reference.

The next time the person phones the call center and gives their name, Voice ID can pull up the voiceprint and compare it to the current voice. If they match up well enough, the call is transferred to an employee who can jump into the reason for the call without having to ask questions about the maiden name of the customer’s mother. If there seems to be a mismatch, the agent will go through the standard screening system to check. The passive test happens quickly enough callers may not even notice that it happened, according to Amazon. The new feature is still undergoing tests but will be a native facet of the platform when it is done. The addition of Voice ID doesn’t preclude customers from using speech recognition technology on the platform from AWS Connect partner companies like Nuance and Auaya, either.

Security Sound

Issues of fraud and identity theft are a constant presence for any business, and call centers are no exception. It’s why many call centers are eager to invest in getting voice identification on their platforms. That’s where companies like ID R&D and their mix of digital authentication technology come in. ID R&D offers a laundry list of voice and visual biometric data collection and analytics. The company has a wide range of options, including many tools that don’t need the cloud to identify someone’s voice.

Voice identification is becoming an important aspect of consumer voice assistants too. Amazon created voice profiles for Alexa users back in 2017 and has been augmenting the feature ever since. This summer, the voice assistant started applying user contact information to personalize interactions with Alexa. Google, too, has been keen to get people to use the voice match feature on Google Assistant. The company completely revamped the setup process for the feature this summer, adding more steps to improve security and make the voice assistant more flexible when personalizing its responses.

The ongoing COVID-19 health crisis has only accelerated the trend toward digital customer service and the need for securing people’s identity for call centers. That’s especially true with more call center employees at home who would normally be in an office this year. Voice ID enables their employers to confirm it is them answering phone calls, just as much as security cameras do visually.


ID R&D Shrinks Voice Biometrics to Internet of Things Edge Processing 

Alexa Developers Can Now Personalize Voice Skills With Names and Phone Numbers

Sensory’s New VoiceHub Platform Offers Quick Custom Wake Word Creation