Key IVR continues to innovate and investigate new technological breakthroughs to determine what features can be added to our services that bring real, tangible benefits to our clients. All part of our mission to deliver the best solutions possible and a superior experience to customers.
AI technology and machine learning have exploded recently in capability and possibilities. With AI models such as ChatGPT and DALL-E being integrated into a range of applications, services, and business operations, it’s inspiring to see how cognitive architecture is being applied. Organizations are using AI to deliver websites, smart devices, and applications that recognize the world around them, generate images and text from simple prompts, analyse huge amounts of data to provide clear insights or identify trends, and much more. Within the payment and CX industry, AI has been leveraged to help combat fraud, communicate with customers as virtual agents or bots, identify payment risks, and protect customers and organizations from malicious activity.
After months of hard work and real-world trials, we’re pleased to announce the first stage in our AI-powered enhancements, our voice recognition service.
Voice Recognition Success Rates Jump, Thanks to AI Technology
Our Automatic Speech Recognition (ASR) engine has developed quickly over the years as we strive to improve the accuracy of how spoken words, letters, and numbers are converted into digital formats. This voice recognition component features in many of our telephony services, such as Payment IVR and Agent Assisted Payments, for speech-to-text audio transcriptions and as an alternative to using a telephone keypad.
We’ve implemented Artificial Intelligence (AI) technology to improve the reliability of the natural language processing of our Automatic Speech Recognition (ASR) engine. The improvement in success rates across our environments has been significant, helping us address any hesitations organizations may have when implementing voice recognition within their payment services and customer communication channels.
When Is AI-Powered ASR Useful?
Our AI-powered ASR engine (also known as Computer Speech Recognition, or Speech to Text (STT)) delivers the most benefit when working with poor-quality or unclear telephone audio. In real-world situations, a customer may be speaking over the phone with a noisy background, have a poor phone signal, speak in a strong regional accent, or much more. If a voice recognition service has low accuracy or long delays, this can lead to a poor customer experience as they’re forced to repeat information and could abandon the call.
In these instances, AI-powered voice recognition can use machine learning to interpret and translate the audio into data with a higher degree of success.
There are two examples where AI-powered ASR could be used:
- Customers on-site provide car registration or VRN details when paying for a car parking space, permit, or Parking Charge Notice (PCN). Letters and numbers combined can prove challenging for other recognition engines to distinguish and accurately identify.
- Customers at home, in a work environment, or outside on a busy street, providing address details for gift aid or a delivery. Building names, streets, and towns can have a wide variety of names comprised of non-standard words. The last thing a customer expects is their delivery going to the wrong location; they may also abandon an order because the phone line can’t identify their address correctly.
The data interpretation and recognition all take place on our PCI-DSS and ISO 27001-accredited platform, with no connection to cloud services or third-party providers.
We will continue to research and investigate new technology, evaluating how it can improve our services for our clients. There are several other AI-related initiatives being developed that we look forward to announcing soon.
Find out more about our AI-powered Automatic Speech Recognition (ASR) engine, or contact us today, and we’d be more than happy to discuss your requirements and arrange a demo.