Pypestream Executive Interview
Richard Smullen, CEO, Pypestream
Click image below to read the ebook
Sheri Greenhaus, Managing Partner of CrmXchange, had a
chance to discuss Pypestream with Richard Smullen, CEO Pypestream, Inc.
Pypestream, a revolutionary platform, aims to
transform customer-business interactions by eliminating the need for middlemen.
Through automation, visualization tools, and AI, Pypestream offers a
self-service model that allows users to perform tasks seamlessly. With real-time
language translation, proactive troubleshooting, and access to customer
records, Pypestream enhances the customer experience. The platform's unique
Pype interface combines app-like functionalities with chat, setting it apart
from other providers. Pypestream's analytics-driven approach ensures continuous
improvement and enables organizations to automate various aspects of their
operations. With partnerships and successful implementations, Pypestream has
proven its capabilities in major enterprises.
Could you provide us with an overview of
Absolutely. We've been in operation for a
few years now, with our headquarters based in New York. Our main objective from
the beginning was to eliminate the need for a middleman in customer-business
interactions. We wanted to revolutionize the way these interactions take place.
The idea came to me around nine years ago, during the time
when Uber was gaining significant traction. At that time, in order to book a
car, you would have to make a phone call to speak with someone who would then
contact the driver. They would send you a form to fill out via email, asking
for your credit card details. You would then receive a call back with the
driver's details and the estimated arrival time. Travis Kalanick and his team
at Uber figured out how to automate this process and make it entirely digital.
Considering the advancements in technology, it's highly
likely that all businesses will eventually adopt self-service models and interact
with customers similarly to how Uber does. However, for this vision to become a
reality, three key elements needed to come together and work seamlessly.
Firstly, automation had to reach a level of excellence. It's
now clear that most tasks can be automated without requiring human
intervention. If processes aren't automated, businesses are essentially wasting
money, and human resources are better utilized elsewhere.
The second element was the advancement of visualization
tools and mobility. Users needed the ability to engage with systems and
platforms from anywhere using their mobile devices. Swiping, typing, and
pin-dropping functionalities had to be highly intuitive and efficient.
Lastly, artificial intelligence (AI) needed to be capable of
understanding human behavior better than call center agents. This was the
missing piece of the puzzle. If AI could replace agents and interact with
humans while taking context into account and analyzing an unlimited amount of
data points, it would significantly enhance the customer experience while
reducing the need for human agents.
Until about six months ago, AI had not reached the necessary
level of accuracy and understanding. The chatbot experience often resulted in
confusion, leading to escalation to human agents. However, we anticipated the
development of large language models and generative AI tools about 18 months
ago, preparing our platform for their integration. This allowed us to deliver
what we call the "call center of the future," where users can perform
tasks and interact digitally without the need for extensive training on
We have been aggressively rolling out this approach and have
already partnered with major companies such as MAX, Sling TV, HBO, T-Mobile,
and the Department of Veterans Affairs. Procter & Gamble is using our
platform across eight of their brands, with global implementation on the rise.
Our proposition is worth taking seriously because it
represents the next generation beyond the failed chatbot era. Users can chat
within the interface, swipe, pin-drop, and our system can proactively
anticipate their needs. This level of seamless interaction is unmatched.
sets Pypestream apart from other providers who claim to offer similar
The main distinction is that we can prove
our capabilities by showing real solutions in action. You can visit companies
like MAX or Clear at the airport, and you will be using Pypestream. When you
ask other vendors to demonstrate a real solution, most of them disappear or
show you a chat window with a human on the other end. We can actually let you
experience our platform and run through various use cases, which immediately
highlights the difference and effectiveness of our approach. We believe in
showing rather than just telling. Additionally, we can back our claims with
partnerships and case studies.
We wouldn't be able to talk about our success with partners
and clients if it couldn't be validated. With thousands of chatbot vendors in
the market, our goal was to create an experience that is unlike any other. We
have developed something called the Pype, which combines the functionalities of
an app and a chat interface into one hybrid experience. It's part of our unique
pattern, and when users see it, they realize it's a lot more functional and
innovative. We often hear that they have never seen anything quite like it.
It sounds like Pypestream has the potential
to automate various aspects of businesses. Could you elaborate on how
automation spreads throughout an organization once the initial problem is
Absolutely. Once we address the initial
problem, our automation capabilities can be extended to various other areas of
the business. We can automate nearly anything within an organization. This is
one of the remarkable aspects of Pypestream. Once we're integrated, we
continuously improve the business through automation, extending its benefits
beyond the initial use case.
Can Pypestream's tools detect someone's
emotions in a panic situation?
Detecting emotions would require access to
data that indicates a panic situation. The key question is whether there is
good access to that data in the given moment. Pypestream analyzes real-time data,
such as location, history, and text construction, to infer the emotions of a
user. Although it's not yet 100% accurate, it's well on its way to becoming
If I provide my identity, can Pypestream
retrieve my record from the customer relationship management (CRM) system and
understand why I may be contacting them?
Retrieving customer records from the CRM
system is a fundamental capability of Pypestream. We examine historical
interactions, remedies provided, and re-contact history. We analyze whether the
customer's inquiry is related to a previous remedy and determine if the
resolution was successful. Our goal is to minimize frustration and provide
proactive solutions. Moreover, in real-time, we can identify issues like
buffering for a streaming provider. Before the customer even contacts the
provider, they receive a text message from Sling TV notifying them of the
buffering issue and offering troubleshooting options. This proactive approach
adds significant value for our customers.
It seems like Pypestream's capabilities are
vast. Where should organizations start when considering Pypestream?
This is a common challenge for executives
right now. We typically begin by analyzing the contact center and understanding
the volume and the use cases driving that volume. We examine the speed at which
we can connect the use cases to backend systems. We prioritize the automation
of use cases that can be done quickly and will immediately impact business
operations by saving costs and reducing the need for human agents. We analyze
which use cases can be automated and how fast we can develop them. Once we
identify the top use cases, we build them out and put them into action. Our
process involves a deep dive into the analysis, which we offer to our customers
free of charge. We want to ensure that our automation delivers real impact and
aligns with the goals of our customers.
Is Pypestream available in multiple
languages, or is it primarily in English?
Pypestream is deeply integrated with AWS,
which provides us with essential core tools. One of those tools is Amazon
Translate. Through our integration with Amazon, including services like Alexa
and Amazon Prime, we have access to an extraordinary translation service within
AWS. At the moment, we support approximately 100 languages at a high level of
When someone is texting, does Pypestream
translate the messages in real-time and store them in a database?
Yes, Pypestream detects the language of the
user's browser and translates messages in real-time accordingly. Whether it's
proactive outreach or responses within the platform, Pypestream leverages
real-time translation capabilities to provide a seamless multilingual
The capabilities of Pypestream seem
limitless. Is there anything else you believe our audience should know about
Analytics play a crucial role in Pypestream.
Our platform is built on a robust analytics layer, which means that every user
movement within our micro apps is analyzed in real-time. We have an extensive
library of micro apps that we can customize for our customers, and we also
build new ones as needed. Once the platform is up and running, continuous
tweaking, fixing, updating, and refining occur to optimize the user experience.
We offer complete visibility into the effectiveness of the platform, leaving no
room for guesswork.
I want to emphasize that we only charge our customers if we
successfully prevent the call from reaching a human agent. Our partnership
model is performance-based, and our pricing reflects a small percentage of the
cost of a traditional call center interaction. The rest, about 90%, becomes
pure profit for our customers. We work closely with each company to define
specific goals, and the impact on their business can be substantial, leading to
improved shareholder perception and even influencing share prices.