Artificial Intelligence

Alacriti CEO: Voice And The Rise Of ‘Elegant’ Payments

“Elegant” as a descriptor doesn’t make many appearances in the world of financial services or payments, which tend to favor other terms. “Disruptive” is a common attribute, as is “innovative.” “Cutting-edge,” “friction-free” and “one-of-a-kind” also hold sway.

That, Alacriti’s Co-Founder and CEO, Manish Gurukula, said, is about to change. The ability to use voice to do something as simple as paying a bill is what will tip the scales.

“In the age of voice-activated UX [user experience] in general,” Gurukula told Karen Webster in a recent conversation, “users expect that a transition to a conversational mode of interacting should mean that ‘technology’ just fills in a lot of the usual information gaps without having to ask the consumer.”

The implication, Gurukula said, is that developers must learn to build skills that understand and can respond to user intent.

It won’t be easy, simple or straightforward, he explained, particularly in a vertical like healthcare, where HIPAA only amplifies the already high requirements for securing and managing sensitive consumer data.

And consumers expect their interactions with those service providers to be channeled or device-agnostic — in other words: Their experience is the same no matter where they’re accessing the skill.

 

Elegance Meets Security

The Alacriti team has spent the last decade and a half developing payments technology that supports bill payment for verticals including financial services (think online bill payment through banking apps),  insurance, healthcare and utilities.

Gurukula said the bill payment landscape has changed significantly over the last 15 years, and even more quickly over the last several, as chat-based interfaces (particularly using voice) have emerged and have provided an opportunity to create an elegant, simple and streamlined way for consumers to pay bills — and an opportunity for development teams to rethink these transaction flows, while keeping consumer data private and secure.

It began with a simple text-based chatbot for use with Facebook’s Messenger service, the CEO said, with the goal of making payment interactions like bill pay simpler. That effort has been in beta testing with several firms for the last year and will get its first full launch next week.

As the Alacriti team was building out that functionality, it became clear the next generation of conversational transactions were going to be carried out via the various emerging voice platforms, which meant they needed to begin evolving their offering along with their ecosystem.

“As we were contemplating how to move our text-based Messenger chatbot capabilities to voice and Google [Home] and Alexa, we had to first answer a few key questions: Where does the user data reside, how is it handled [and] how is the flow from one endpoint to another managed?” Gurukula said.

Those foundational questions, he said, can be answered in a few different ways in order to connect an elegant experience to a secure one. Alacriti, for example, does not allow users to add payment information directly through a voice device or Facebook Messenger. Instead, that information is collected during the user’s registration process.

The key, Gurukula noted, is to leverage the best practices that PCI, and HIPAA in the case of healthcare, have already put in place in order to minimize data exposure to what is necessary for the transaction workflow to move forward.

 

The One-Time Payer

Many consumers have automated recurring bill payments to flow directly from their bank accounts. Maybe they get a text or an email alert that an automatic payment is scheduled, but most autopay customers like the autopilot aspect of autopay.

However, Gurukula noted, there are many one-time payers that exist in a variety of contexts.

In healthcare, for example, people encounter all kinds of players in a one-time-only capacity — specialists, for example, or doctors visited for a single consultation.

The problem, Gurukula said, is that when it comes time to pay, the customer has to deal with whatever payment mechanism the provider has set up, and those provider portals can require a level of information provision that rivals trying to get security clearance from the federal government. Those very robust, secure portals make sense in the context of a patient with a long, ongoing relationship with a provider. where there are all kinds of sensitive data points to be managed.

But for a one-time visitor, it’s a bug, not a feature — and a bug that Alacriti can fix by offering a more streamlined — but still compliant — one-time payment platform.

In the wider world of recurring bill payments — for utilities, credit cards and the like — one-time customers are likely not one-off customers. These customers show up once a month to pay a bill, but (for whatever reason) they don’t want to set the bill to automatically pay out of their bank account.

“That is a fairly large number of customers, and that is the segment of the market we are serving,” Gurukula told Webster.

The goal for that customer is to make it as easy for them to say out loud, once, “Alexa, I would like to pay my bill” and be done with that process in one or two quick, simple conversation interactions.

“More than that,” he noted, “the process gets tedious and annoying quickly,” immediately squandering the advantage of voice by irritating the customer when it should instead be delighting them.

 

A Real Artificial Assistant

The potential inherent in Amazon’s Alexa or Google’s Home Assistant, according to Gurukula, is to offer a better and more elegant way to interact with the elements of an individual’s day-to-day life.

The voice assistant, he noted, does better than just responding appropriately to its user’s requests; it comes to understand what the user is requesting, so it can fill in more of the steps on its own and keep its user moving forward, without having to think much about the instructions it’s giving an assistant.

“This means that when the customer says, ‘Alexa, I want to pay my bill,’ or ‘I want to pay my bill now’ or ‘I want to settle up,’ the skill we have programmed is strong enough to differentiate what it is hearing in a few consumer contexts — and still serve up the right action,” Gurukula said.

Because a good assistant is one that “needs as little instruction as possible. Otherwise, you’re assisting it.”

That means, for example, the virtual assistant should “remember” the customer and what his or her previous habits were like. When a customer says she wants to pay her bill, the assistant knows to ask if she wants to use the same method she used last time to pay and can then complete the transaction when it gets confirmation.

Or it means the virtual assistant can handle a slightly more complicated command, so that when the customer says, “Alexa, I want to pay my power bill with my Discover Card ending in 1234,” it can process both of those ideas at the same time and get the job done.

Or it can even mean, Gurukula said, greeting the user in the morning, noting the daily rundown of payments due and offering the option to settle up in the typical way while the customer is in the shower getting ready in the morning.

At this point, he told Webster, Alexa isn’t just a repository for a bunch of digital skills, she’s literally the user’s assistant programmed at every corner to make life a bit smoother, a bit more efficient and a bit more friction-free.

Or, in a word: a bit more elegant.

Alacriti is hoping to help push Alexa’s potential forward with the skill it’s prepping for the PYMNTS Voice Challenge With Amazon Alexa. All submissions will be on PYMNTS for public voting. The top five submissions will get to pitch their skills at our Annual Awards Ceremony in April 2018.

We look forward to finding out if Alacriti makes the cut.

——————————–

Latest Insights:

Our data and analytics team has developed a number of creative methodologies and frameworks that measure and benchmark the innovation that’s reshaping the payments and commerce ecosystem. In the November 2019 AML/KYC Report, Zillow’s Justin Farris tells PYMNTS how the platform incorporates stringent authentication without making the onboarding and buying experiences too complex.

TRENDING RIGHT NOW