A world beyond chatbots

A spotlight on Pypestream and conversational artificial intelligence

Imagine a world where paying your bills, scheduling a doctor’s appointment, booking accommodations for your business trip, or staying up-to-date on your insurance payments felt as simple and easy as a conversation with a friend. That’s the idea behind Pypestream, a technology company offering an end-to-end conversational interface built for scale and simplicity.

“We think of ourselves as a client experience company more than a technology company,” explains Donna Peeples, President & Chief Customer Officer at Pypestream. At the center of that experience is the notion that transactions which might have left you waiting on hold, rummaging through a company’s help center or scouring a counterintuitive web navigation bar before can now be handled quickly and adeptly by a conversational interface—one who can handle thousands of conversations at once, but through clever design and advanced natural language processing, can make you feel like you’re the only person they’re chatting with.


Nothing quite like it


Artificial intelligence has been steadily gaining speed since the 1990s, but in the last several years, the range of things that can be automated has soared to new heights.1 Pypestream got its start in 2015, when entrepreneur and technologist Richard Smullen decided to create a full-stack, end-to-end solution connecting customer and business through conversation. He engaged a team of developers, designers, and business strategists. It wasn’t long before the startup secured the first of their five patents, and began to catch the eye of customer experience industry leaders like Donna.

Donna, who came to Pypestream with 30+ years of leadership experience in insurance and customer engagement, had been leading efforts to digitize clients’ insurance interactions within an insurance provider when she saw her first Pypestream demo. Right away, she was convinced the team was on to something. “I had never seen anything like it before,” she recalls. “I thought to myself, ‘This is something that could really change the world.’”

Donna decided to join the Pypestream team that week. Four days later, she had moved from Atlanta to a new apartment near Pypestream’s Chelsea, New York headquarters—a modern space complete with an in-office soccer field, a speakeasy, and an energetic office dog named Jagger. In true startup form, it was a decision made at the speed of now.


The difference between a buzzword and a game-changer


Much has been written about Industry 4.0 and how automation is changing our homes, cars and how we interact. And within the realm of AI-assisted interactions, chat is stealing the spotlight as the B2C communication style of choice. “We called 2016 the year of the chatbot,” laughs Donna. Gartner and Juniper now predict that by next year, 85% of our interactions with businesses will occur through AI—and the average person will be having more conversations with bots than with their spouse.2,3 “Children of the screen and people of the page are adopting it more quickly than anyone has calculated,” she adds.

This buzz of activity has led companies hurrying to fit messaging automation somewhere into their model. Bots are popping up everywhere—but, as Donna explains, simply using AI doesn’t make a company’s interactions intrinsically better or easier—it’s how an interface is designed and deployed that matters. “There is a big difference between a simple chatbot and a conversational interface that’s genuinely intuitive,” she explains, “What we do is a world beyond chatbots.”

To paint the picture, she contrasts a messenger that handles a few simple questions with an enterprise-grade interface that includes encryption, military-grade security, and the ability to send rich assets including videos, photos and documents back and forth from AI to user. It’s the kind of tool that can help you book your next vacation, pay your bill, or schedule your next doctor’s appointment—and make you wonder why you haven’t been doing it this way all along.

Just as important as functionality is purpose. AI should be designed, first and foremost, to solve a business need, Pypestream emphasizes. Companies can start with the end in mind by asking two simple questions:

  • What results are we looking to achieve?
  • How do our clients want to interact with us?

Pypestream uses a journey called PypeProSM, designed by Chief Business Officer Evan Kohn, to help businesses figure out where AI will have the greatest impact on their business. “We aren’t having technology people solve business problems,” explains Donna, “We’re having business people solve business problems.” By starting with a clear organizational aim for AI, businesses can discover lower error rates and greater employee satisfaction, not to mention better client engagement.


The making of a great conversationalist


We’ve all heard the words of poet and activist Dr. Maya Angelou, who famously spoke about how people will forget what you say and do, but never how you make them feel. But does her advice still apply when words, actions and experiences are delivered not by a person, but a bot?

Donna believes so. After all, she’s had decades of experience improving people’s day-to-day interactions as they try to navigate the complex—and often confusing—world of insurance. And while insurance may not be clients’ main focus, she explains, insurance providers have the distinct privilege of being there for people at times where they may not be at their best.

“Whether it’s a fender bender, a natural disaster, or even just a payment issue, moments like these can be the worst in someone’s day or week,” she explains. Here, using technology in a way that’s compassionate and forward-facing can make an honest difference. “Even if we can’t change what happened, there’s an opportunity to make it just a little bit better—easier, faster, or more accurate,” she adds.

People love speed, and they love efficiency. We saw this firsthand at Berkley One when we did a listening deep dive into to some of our clients’ and agents’ biggest insurance challenges. Billing was a common refrain, a part of the process which at many companies is decidedly unglamorous and frustrating, especially for new clients. So our CX team started by simplifying the process to a more streamlined, “one-and-done” approach—and then added AI to make it even easier.

Together with the Pypestream team, we created an AI-solution called Bob, a digital insurance assistant who can proactively reach out to clients, allowing them to communicate on their time and terms. He can help them view recent bills, track past payments, and make a secure, encrypted payment in less than one minute, helping them avoid late payments, fees or gaps in coverage with just a few clicks.

At the heart of it all is an interaction that—despite being automated—can feel much more human-friendly. Experiences like this are just the beginning, explains Donna, since AI capabilities such as natural language understanding are getting stronger every day. “AI is changing the way we communicate,” she adds. “Insurers are moving from a model of ‘receive, respond and repair’ to being able to predict, prevent and protect. We’re moving beyond purchases and annual premiums to a continuous engagement cycle.”