Subscribe to get the latest stories
September 10, 2020
Machines, AI, & Future of Customer Service — Our Replicant Series A
Enterprise
  • Scott Beechuk
    Scott Beechuk
Share
Subscribe to blog

I realized something recently when I was picking my friend’s brain about writing this particular post. I shared what I thought was a compelling outline for the basis of the piece, hitting upon the rationale for why I led the Series A investment in Replicant, a company that has developed a product and technology that replicates the behavior of Tier 1 support agents by using advanced artificial intelligence.

It seemed like she was giving considerable thought to my salient points, but after 10 seconds of silence all she said was, “Rutger Hauer and Jude Law can customer service the heck out of me any day.” 

At first, I missed the obvious Blade Runner and A.I. movie references, which was embarrassing considering that’s how Replicant chose its name. Her response got me thinking about the relationship people have with AI and how that relates to the customer service space, especially since AI is where the industry is headed. Sometimes we forget that customer service is not all about resolving a ticket; it’s really first and foremost about listening…which is very “human”. So, when that customer service space is a standalone phone call, it’s even more crucial that whoever/whatever is on the other end of the line better be adept at getting to the heart of the issue. My friend’s response reminded me that we have a very human tendency to assign such traits to things non-human such as AI; in that sense, we are eager to engage as much as the AI can keep up. 

This might seem like a roundabout way of introducing Replicant, but their attention to this humanscale connection is what sets them apart in the customer service arena: they start with the experience between caller and agent and build their AI-driven product around it. The result is a first-of-its-kind product that honors the human aspect of customer service no matter how automated operations get. 

A 360 View of CS

Someone I greatly admire in this space is Tony Hsieh, CEO of Zappos.com. He didn’t just create an online retail empire, he built one of the most respected service companies in the world based on his understanding of people – not just customers but also his agents. He cared about their experience and believed in creating a value-driven culture, empowered them to resolve issues, and prioritized outcomes over metrics. By giving his agents the tools to make a difference, they were more happy and motivated to create that positive outcome. I believe this is the right approach, but it’s dependent on the ability to respond to the volume which, historically, requires people. So that customers aren’t on hold for 10 minutes. So they aren’t cranky at the agents by the time their call is picked up. So agents aren’t rushed or harried. So agents have the ability to respond in an authentic, unhurried, and results-oriented manner. 

There was a time not so long ago when the best and most cost-effective way to provide this level of service was to create a massive contact center with hundreds or thousands of reps sitting in cubicles answering phones and responding to customer requests — that’s how this $25B market was created. The industry then took a leap when digital transformation arrived, building in necessary automations so customer service reps can better respond to the high throughput over many channels at once. Because of the limits of technology, sometimes these efficiencies come at the expense of human engagement. 

Also, while this hopefully is just a blip in our timeline, another thing no one saw coming is COVID and the effect on the industry. Contact centers are operating at a tiny fraction of their capacity and we’re seeing customer satisfaction decrease due to extra hold times with fewer agents. In turn, agents are still doing the same repetitive job, yet are more stressed and less engaged, resulting in unhappy customers and high turnover. It is no surprise that contact center execs are looking for new ways to automate and make up for the loss of human agents. 

Talk Service to Me

Replicant spent three years building the world’s most innovative voice-based conversational AI engine. Replicant’s AI Agent mimics Tier 1 human agents, capable of having long-running conversations and solving customer issues on a single call. 

To be clear, this isn’t solely about having an agent that sounds less like Soundwave and more like Megatron (most of you know what I’m talking about). The innovation here is that Replicant can hold conversations with people more like real humans do — it can remember what you were talking about 10 minutes ago and change topics at any point in the conversation, returning to a previous topic in a natural way. Replicant’s AI brain understands context as well as complex and conversational speech patterns including slang, accents, and humor. (I tested it out. Their AI agent doesn’t think I’m funny either.)

The AI Agent is fast — it holds a conversation in real-time with less than a second of delay with each response, similar to a live person. This allows Replicant’s AI service agents to solve problems for customers in a more human, natural manner than any prior conversational AI system has ever been capable of. 

And, it’s not just about the conversation engine. Replicant is the first voice AI platform to be fully integrated with CRM and contact center software allowing it to authenticate callers, escalate calls as needed, and autonomously capture call notes. In addition to standard service issue handling, customers use Replicant for appointment scheduling, account management, billing management, and insurance policy updates. 

Replicant currently handles over two million customer support calls a month and serves some of the largest call centers in the country. When their AI Agent can’t solve an issue, the system will transfer to the best human agent using an enterprise-grade routing system, reducing the need for multiple transfers. The AI agent hands over the entire conversation history to the human agent so the customer doesn’t have to repeat anything. This means less compromise between efficiency and engagement, while offering a more seamless experience on the customer’s end. 

The speed and accuracy of Replicant’s conversations, combined with its CRM and contact center integration, allow call centers to resolve customer service issues in half the time with 50-75% in cost savings.

Why can’t Interactive Voice Response (IVR) systems solve this problem?

  • IVR’s are designed primarily to deflect calls. Replicant was built from the ground-up to resolve complex issues.
  • IVR’s are passive, single decision-tree solutions, unable to have human conversations with customers. Replicant’s AI Agent can handle customers changing topics in mid-stream or discussing multiple topics at once.
  • IVR’s tend to be slow, with responses greater than 3-5 seconds after the customer speaks. Replicant responds in less than a second for more natural-sounding conversations that also recognize customer intent.
  • IVR’s are static. Replicant is as smart as your smartest Tier 1 agent, and, through machine learning, continues to get better over time. 

Behind Every Good Bot

Replicant is led by Gadi Shamia, a customer service expert from Talkdesk. We got to know each other during my days at Salesforce and I was always impressed with his knowledge of the industry, empathy for customers’ experience, and creative, energetic approach to solving complex technical challenges. CTO and co-founder, Benjamin Gleitz, possesses a remarkable background in AI and ML, and a thoughtful, entrepreneurial drive for excellence. I was already an investor in Replicant’s seed round, but I was so excited about this team and the company’s growth and potential that I doubled down and led the A round.

Humanizing Technology

The dichotomy of what drives the high volume contact center customer service world is really fascinating to me. I don’t know of any other industry that is so intertwined with tech, for its own survival and functionality as well as in service of it. Yet, it has remained stagnant relative to other sectors and is ingrained by what is arguably the opposite of tech – human connection, listening, respect, all subjective and emotional. But it’s just the nature of the business, the coming together of two worlds. Even today’s customers demand these seemingly disparate things: fast yet personal, customized service yet quick fix. 

Maybe that’s what I appreciate about Replicant’s product – it’s scalable, personal, cost-effective, and gets the job done. It helps companies strike a balance that is appropriate for their individual organizations. Brands get a cost-effective way to scale their Tier 1 support that continuously learns and gets smarter as a group unit, and customers get their calls answered instantly and issues resolved quickly. 

For the most complex and technical issues, human agents are still the best route and will be for the foreseeable future. But for more common Tier 1 support issues, Replicant’s conversational AI can reduce the stress on contact centers, especially during times of heavy or spike volume, while providing the speed and accuracy of support that today and tomorrow’s customers expect. Roy Batty not included. 

Subscribe to blog