Image Credit: Getty Images
The first generation of phone banking was largely driven by interactive voice response (IVR) technology. That’s the touch tone-driven technology that provides the monotonous voice tone telling you to “push 3 for your bank balance.” IVR is a technology that was never particularly loved by anyone but it has done the job for many banks around the world for decades, albeit in a suboptimal approach.
A new, more modern approach is now starting to emerge with capabilities such as voice-print authorization, which provides a fingerprint-like feature that enables a user’s voice to authenticate into the system. The droning machine voice of IVR and the frustrating experience of pushing buttons to get through multiple menus is also starting to come to an end with the help of an AI-powered customer contact center platform.
WaFd, formerly known as Washington Federal, is one such financial institution that is now using the Amazon Lex conversational AI technology to help revolutionize how it does phone banking. WaFd is based in Seattle, Washington, and has more than 200 branches across eight states. The company has built out its own digital enablement organization, known as Pike Street Labs, to help drive technology initiatives forward.
Intelligent Security Summit
Learn the critical role of AI & ML in cybersecurity and industry specific case studies on December 8. Register for your free pass today.
For WaFd, moving to conversational AI was also a way to help improve user experience by helping customers get the information they want faster. Dustin Hubbard, chief technology officer at WaFd Bank and Pike Street Labs, told VentureBeat that previously it took about four and a half minutes before a customer got to the point where they could press a button to actually get their balance.
“Now when you call in, the system knows if you are voice authenticated, which means you can prove your identity,” Hubbard said. “You say, ‘My voice is my password’ and the system responds ‘Great, how can I help you?’ and at that point, you’re having a conversation with the chatbot directly.”
Instead of the user needing to go through the IVR menu to get to the right spot to be able to find out what button to press to get what they are looking for, the customer just has to ask for what they want. Hubbard estimated that instead of four and a half minutes for a user to get a bank balance, they can now get it in approximately 28 seconds.
Several components of the WaFd platform have replaced the legacy IVR system.
The voice-print authorization capability comes from Talkdesk, which has a cloud-based call center as a service offering. Hubbard explained that when a customer calls the WaFd banking number, the call gets picked up by the call center system. The voice authentication system verifies a user’s voice, as well as analyzing the phone number and location from which a call is coming in, before granting access to an account. The call center system can then log into the WaFd backend online banking system through a series of APIs.
Once logged in, the Amazon Lex conversational AI technology kicks in. Amazon Lex is the foundational conversational AI technology behind Amazon’s Alexa voice assistant. Amazon’s Lex service knows that the user is authenticated and will let users simply ask to check a balance or to make a transfer across accounts.
Even though WaFd is using two systems, with Talkdesk and Amazon Lex, Hubbard said it’s entirely transparent to users who don’t know they are transitioning between systems. In order to enable the seamless integration, especially in terms of the actual voice that users will hear, WaFd is using Amazon Polly technology.
Amazon Polly provides a text-to-speech capability that WaFd has used to record the voice prompts that the user will hear in the Talkdesk system when they initially call in. WaFd is using the same voice with Amazon Polly that it is using with Amazon Lex. Hubbard said that he wanted to make sure users got a consistent experience with the same voice.
Show me the money — how Amazon Lex uses utterances to train conversational AI
Getting Amazon Lex trained for the WaFd banking use case didn’t involve a massive amount of effort.
Hubbard explained that Amazon Lex is trained with an approach known as utterances. For example, by giving the system the phrase “check balance” and variations that users might use in a normal call, the system is directed to check the user’s account balance.
Over time, Amazon Lex gets better at inferring what the user intent is. For example, if a user doesn’t exactly say what the trained utterance is, the system is able to infer the probability that it’s close, and it will still execute the expected action.
“That’s how the conversational AI part with Amazon Lex is better than a lot of other systems that I’ve seen, where if it’s not an exact match, the thing just basically has no idea what you’re talking about,” Hubbard said. “You could say, ‘show me the money,’ for example, and the system will know that you want to check your balance. And generally, over time, it gets better at recognizing patterns.”
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.