Now, the next generation of AI tools like ChatGPT and Large Language Models (LLMs) are helping us take our vision further. 

Leveraging data gathered from hundreds of millions of customer interactions, we harnessed the power of LLMs to get our first customer live and in production using LLMs and delivering 90% resolution rates. 

Our journey has always been about putting customers first by using automation to solve their problems instead of merely deflecting their calls.

This unique approach has resulted in tens of millions of satisfied callers who’ve had their issues resolved swiftly and efficiently, with no wait times or endless transfers.

Now, LLMs and ChatGPT are taking our solution to new heights.

These groundbreaking AI technologies come equipped with an extensive knowledge of the world, which make them perfect tools to help automate a wider range of customer inquiries.

When paired with our comprehensive Contact Center Automation platform featuring high-availability telephony, omnichannel, out-of-box integrations, flow design, analytics, A/B testing and more, LLMs become transformational tools to help tackle customer issues with greater success and flexibility.

We’re proud to be among the first to deploy LLMs in live customer calls in the contact center.

Our team’s deep experience with language models allows us to move quickly. And, because our Thinking Machine is already designed with enterprise-grade security and scale to ensure customer data is always protected, we are able to instantly ingest LLM data with the same protections. 

With a faster, more accurate and knowledgeable Thinking Machine, we now see fewer caller escalations and enable contact centers to resolve more customer issues more efficiently, and with shorter setup times.

So, how does this benefit our Contact Center Automation platform?

• Resolve more customer issues. Higher resolution rates and lower handle times result in better customer experiences and cost savings.
• Faster time to value. More flexible conversations delivered faster make it easier than ever to reap the benefits of automation.
• Safeguards and controls. Robust guardrails that maintain the highest standards of reliability and security. Every Thinking Machine remains SOC2, HIPAA and PCI compliant, and PII or sensitive data is never passed to third-party LLMs

The 2023 Benchmark Report highlights the importance of LLMs for Contact Center Automation, with 93% of contact center leaders preparing for an economic downturn and 68% prioritizing automation as their top technology investment.

Our partnership with CAA Auto Club, the leading roadside service provider in Canada, showcases the incredible impact of LLMs. Our Thinking Machine already cut emergency roadside calls’ handle times in half, and now with LLMs live, we have improved call completion rates and allowed CAA’s members to enjoy a more flexible and natural experience during emergency roadside service calls.

As AI continues to evolve at a rapid pace, we remain committed to harnessing the power of LLMs without compromising security, redundancy, and compliance. Our exceptional team’s hard work and creativity have been instrumental in achieving this milestone, and we couldn’t be more excited about the future of Contact Center Automation.

So, buckle up and join us on this exhilarating ride into the future of customer service!

The post How Replicant Unlocks the Future of Contact Center Automation with ChatGPT and LLMs appeared first on Replicant.

In fact, many of the issues we’re seeing today started well before the lockdowns. Government attempts to keep people and businesses afloat during 2020-2021 masked these warning signs, contributing to the sense that things may get worse before they get better.

At Replicant, we’ve been exploring the impact of this anticipated economic downturn on customer service and have produced a variety of helpful resources for contact center leaders to consider when analyzing the challenges – and more importantly, the opportunities – that lie ahead.

Experienced customer service professionals understand the importance of investing in technology that enhances productivity, reduces costs and promotes customer loyalty. And this priority isn’t slowing down as you’d expect in a recession.

Spending on enterprise software is actually expected to continue to grow as leaders identify crucial “must-have” pieces of technology that can both cut expenses and improve the customer experience. 

An analysis from Gartner projected global software spending to hit nearly $675 billion in 2022 – up 9.8% from a year earlier. This momentum is expected to continue and surge another 11.8% in 2023 to almost $755 billion.

“CIOs are accelerating IT investments as they recognize the importance of flexibility and agility in responding to disruption,” said John-David Lovelock, research vice president at Gartner. Clearly, there is a massive opportunity here. As the Gartner research indicates, tools that directly contribute to productivity will be prized by innovative IT professionals.

In light of increased IT spending amid changing economic dynamics, automation has rapidly become a critical resource for contact centers and will continue to gain prominence as leaders look to improve customer satisfaction, backfill missing agents and reduce costs.

I recently wrote in Forbes about how customer service is in a state of crisis, driven by an outdated model for contact centers relying on human agents forced to do repetitive, monotonous work. It is impossible to predict the number of agents necessary at any given hour, leading to understaffed operations – and furious customers. A highly scalable service, contact center automation removes the staffing headaches we’re seeing across so many industries.

But contact center automation isn’t about replacing the role of human agents. It’s about automating the tedious aspects of the job that become tiresome and lead to such a high level of turnover. And don’t take my word for it: a large utilities company in the midwest told us that they’ve seen up to 20% agent attrition a month.

A transportation company is still unable to hire enough agents just to get back to pre-pandemic numbers, which is about 30% short of its target. In a hybrid model, agents are instead empowered to work with customers who have more nuanced and complex requests, improving the quality of their job by giving them meaningful work, and providing customers with a more engaged support system. 

So what does this mean for customer service and contact center leaders like yourself? We’re seeing the impact of increasing investments in “must-have” technologies firsthand with Replicant customers.

One of our longtime clients, the Canadian Automobile Association (CAA), adopted our contact center automation platform to be “prepared for whatever may come,” according to Steve Bennett, supervisor of member care for CAA. “Whether it’s an economic downturn or a spike in call volume tomorrow, we will be able to answer calls because of the automation we already have in place.” 

While automation can’t solve all problems, it can help businesses spend less time and resources resolving mundane issues. This allows agents to focus on more complex and rewarding customer interactions, as well as increases the quality of customer service – therefore reducing customer churn and ensuring satisfaction with a brand.

The post Swimming Against the Current: Empowering Your Business During an Economic Downturn appeared first on Replicant.

It is never fun to be locked out of the car, but this day was particularly bad. As a loyal CAA (Canadian AAA) customer, Noah knew he would get help, but in those conditions, even a few minutes on hold would be brutal. Like in the US, the Great Resignation has wreaked havoc on Canadian call centers, and waiting for hours on hold became the rule, not the exception.

But when Noah called CAA, the phone didn’t even ring once before the call was answered, and a natural-sounding bot called a Thinking Machine answered.

Recognizing Noah by his phone number, it was quickly able to find his location, understand the problem, collect all the needed information, and send a locksmith his way in less than four minutes. 29 minutes later, help arrived and Noah was on his way.

The Thinking Machine managed 2,834 similar calls that day. 

We all dread calling customer service- now more than ever. Will it take 20 minutes? Will it take an hour? Will it take five hours? Navigating through options that “may have changed,” listening to marketing messages on a loop, and finally speaking with an agent calling in from somewhere on the other side of the world and in the middle of their night. While it is easy to be angry with the agents, it is hardly their fault.

They have to deal with unexpected call spikes due to the pandemic and supply chain issues, manage higher call volumes, and deal with angry customers. It is no wonder that over 50% of them reported taking anti-anxiety medications, and we can all assume the pandemic made it much worse. While companies do their very best to deal with higher call volumes and spikes, not enough agents want this work, and the ones who join often leave within months. 

If this seems like a huge problem to you, we are on the same page. This is why we started Replicant, and this is why some of the best consumer brands in the world are adopting our product. For 30 million callers, wait time was not an issue, and their problems were solved in the time it took to refill their coffee. With an 85% success rate, dozens of different types of calls that are now fully automated, and always-on availability, we are making a real impact on our customers’ business and reputation, but more importantly, on their customers as well. 

To help us keep building the future of customer service, I’m excited to share that we’ve raised a $78M Series B round to help scale our platform and team. This round will bring our total funding to $113M, will support our growth and help our customers automate more of their customer service flows across all channels.

New York based Stripes led this round, and Ron Shah will join Replicant’s board. Stripes has invested in prominent SaaS companies like Monday.com and category creators like Udemy and Fullstory. Also joining this round are Salesforce Ventures, IronGrey, Omega Venture Partners, and Alumni Ventures. Our existing investors, Norwest and Atomic, also participated in the round. 

All of this momentum would not have been possible without the incredible employees of Replicant. I want to thank the Replicant team members- so many of you joined us without even a chance to meet in person. 

Also, my eternal gratitude to our first customers who believed in us and were willing to give us a chance- knowing that they were investing in powerful, but still early, technology. Those visionary customers are often the unsung heroes behind the success of so many SaaS companies. 

We are focused on developing our Contact Center Automation platform and helping more companies achieve customer service bliss. We plan to invest in expanding our platform and hone our conversations so we can resolve even more calls and service issues. We will continue to invest in our team to help them grow professionally and achieve their career goals.

Everything else falls in place when people like what they do and who they work with. We are in the early days of Contact Center Automation, and we could not be more excited for the years to come.

Thank you!

Gadi Shamia

The post Building the Contact Center Automation Category: Replicant’s $78M Series B appeared first on Replicant.

During the decade preceding 2020, customer service ran predictably. Most companies grew steadily and leveraged the spread of fast internet to outsource customer service to Europe, then to Manila, and recently, to smaller towns worldwide. Efficiency was gained through a less expensive workforce, and the challenges of hiring, training, scheduling and maintaining adherence, were outsourced as well. While these strategies didn’t solve the inherent challenges of operating large scale customer service organizations, they pushed them farther out of view.

At the same time, a new technology started to emerge – Artificial intelligence (AI).

A vague promise at the start of the decade, AI is finally taking shape and starting to make a real impact. With an understanding that AI could resolve some of the fundamental problems in customer service and have meaningful, human-like conversations with customers, we brought Replicant to market.

As covid-19 sent a shockwave throughout the world, companies realized that they could no longer rely on human-only service models. Agents, especially international agents who could no longer take phone calls from crowded homes in the middle of the night needed help to support countries in different time zones. That was Replicant’s time to react. For one customer, we launched a Thinking Machine in three days that answered all of their incoming customer service calls, triaged and prioritized them, and assigned them to the right technician. We launched Replicant quickly so their customers – fast-food restaurants all over the country – could continue operating their equipment at scale. For another customer, we automated a major digital business process in their supply chain, taking on 30,000 complex calls per day in less than six weeks. This customer was able to save nearly $4m a year on their call center operation costs and improve customer satisfaction by reducing order errors after implementing Replicant.

A lot had to happen in 2020 to support an uptick in customer demand for Replicant. We started the year with fewer than 20 employees and finished with 55. A 500X increase  in call volume sent the engineering team to add more servers, launch two high availability data centers and polish their code. While not pausing for a minute, we raised $27M in series A financing from Norwest Venture Partners, adding one of the key executives behind Salesforce’s $5B Service Cloud business to our board. And, we did all of this while staying true to our values: 

We only engaged with customers when we could take on calls we could resolve, rather than route or deflect.

We stayed away from enabling unwanted outbound calls and telemarketing, always keeping the customer on the other end of the line front and center. 

We ensured that every customer we signed-up, went live and achieved their business goals, often after trying and failing to use well-known players. 

I am very proud of the team’s work and hope you demand the same standards from us as we move forward together. 

As we start 2021, we hear a constant theme from customers and soon to be ones: elasticity is our number one priority. With unpredictable vaccination campaigns, new strains of viruses, and ever-changing local restrictions, no customer service leader can predict customer demand and expected call volumes. If they could have waved a magic wand, they would have preferred to start 2021 with flexible capacity that can expand and contract in minutes according to customer demand. When you harness AI to design an elastic, Tier-1 first line of defense for your highest volume calls, you can protect your most experienced agents from fluctuations in demand, offer faster customer service, and reduce costs as you no longer have to pay for unused capacity, and can even save on per-minute costs, as AI is cheaper to operate. The promise of Replicant’s elastic customer service is exciting not just for the customer experience but for the industry as a whole.

And, as we’ve seen in the past, major market shifts always boost automation. A few years from now, interacting with a Thinking Machine on the phone or via text will be as natural as interacting with a live agent. And just like ATMs did not replace humans but allowed them to instead focus on more complex tasks, Thinking Machines will team up with agents to provide the most optimal customer service imaginable.

We all hope to finish 2021 better than we started it – vaccinated, hosting our friends and family again, and enjoying travel. We can also end 2021 with a new customer service strategy that will not only help during these testing times but will serve you and your customers for decades to come. 

Sincerely,
Gadi Shamia

The post Looking Back, Sprinting Forward: Replicant’s 2020 Year in Review appeared first on Replicant.

This investment came after a few quarters of rapid growth, scaling from answering thousands of calls per month to millions in less than two quarters, following an accelerated demand for automation in customer service. During this time we doubled our team, released dozens of new features, and helped Fortune 500 customers get up and running in days, helping them to maintain their customer service operations in difficult times.

Covid-19 brought a new set of challenges and impacted every industry overnight – take e-commerce for example – customers could no longer go in-store to make purchases or interact with store reps and turned to online shopping. With e-commerce purchases doubling in volume, demands on customer service departments spiked accordingly. Add to that unpredictable agent availability caused by lockdowns and shelter in place orders, and the result was long wait times, overworked agents, and unsatisfied customers. The pandemic exposed the major shortcomings of call centers – relying on scheduling enough trained agents in physical locations to answer an often unpredictable and spiky volume of customer calls.

I knew it was time to address the bigger challenges facing customer service and fortunately, my journey in this industry started earlier on in my career. In my twenties, I ran a call center and many years later joined Talkdesk, now a $3B contact center software market leader, as their first COO. With this deep expertise and passion for customer service, the next step was clear – helping lead the innovative team at Replicant to harness the power of AI to make calling customer service as seamless as ordering an uber or making an online purchase. In contact center speak, this means delivering an effortless customer service that eliminates hold times, resolves customer issues quickly, increases operational efficiency, and improves agents’ jobs. 

With this vision in mind, Replicant.ai was born – the world’s first autonomous contact center that fully resolves tier-1 customer service issues over the phone using the power of voice AI. With Replicant, companies can provide 24/7 service, eliminate hold times, and scale elastically to meet high call volumes while improving customer satisfaction and reducing costs. Beyond resolving customer issues, Replicant transcribes every word that’s spoken by customers, giving companies deeper insights into customer data. It integrates into existing call center software to authenticate callers, escalate calls with context when needed, and captures call summary notes autonomously. 

We have seen incredible success with our initial customers, including some of the country’s largest call centers, and are honored by the trust they’ve placed with us. Replicant customers are reducing costs by 50%-75%, cutting average handle times in half, resolving more than 90% of calls without escalations, and improving customer satisfaction as a result of shorter and more effective calls. We recognize the responsibility of serving our Fortune 500 customers’ 24/7 operations and have invested in scalable data centers, state-of-the-art security best practices, and strict reliability and privacy certifications like SOC2 type 2, HIPAA, and PCI to achieve the highest levels of trust and security.

As we continue our journey forward, our vision for Replicant’s autonomous contact center is to become the call centers’ first line of defense so that agents can focus on what they do best – fostering relationships and championing brands. Following our latest financing round, we plan to triple our investment in R&D, grow our customer support team, and expand sales and marketing to deliver seamless customer service fueled by AI and automation. Today’s milestone is another step in our journey to transforming the future of customer service for customers, agents, and brands. 

And, as more and more of our experiences become digital and move online, customer service is becoming one of the most critical interfaces between brands and their customers. Customers that can hail a ride or book a flight in a single click, want the same convenience and speed when they need help. Replicant’s technology is making what used to be science fiction not too long ago, into a reality: having a Thinking Machine answer customer calls without hold times and resolving customer issues politely and effectively in minutes. It is time to bring customer service to the twenty-first century, and Replicant is doing it now. 

The post Replicant.ai raises $27M to modernize customer service with the world’s first autonomous contact center appeared first on Replicant.

Contact centers are accustomed to spikes in demand for service. Tornados, hurricanes, snowstorms, the big holiday shopping days—are all short-lived situations that the best contact centers master with a minimal or short-term impact on customer service. But COVID-19 sets a new bar for just how nimble contact centers need to be to ensure great customer service. Around the world, contact centers have temporarily closed, lost agent capacity, or are struggling to ramp up remote work. As a result, many consumers face long hold times—if they even get that far.

The airlines provided a telling example. When COVID-19 hit, call volume was up as customers canceled or rescheduled flights with fewer agents coming to work due to exposure risk or local lockdowns. More demand, less supply, and the result was tweets like this.

Some companies are fantastic at handling spikes in demand. Uber, for instance, rolls out higher prices during surge traffic times to lure more drivers into service. When the surge subsides, so does the higher pay. Driver supply naturally falls and equilibrium is again reached.

But contact center agents are not drivers. Driving a car is a generic skill that doesn’t require specialization. Google Maps gives drivers the ability to navigate any city regardless of whether they’ve been there before. Agents are highly specialized and require company-specific training. They need access to internal systems, constant updates on protocols and scripts, and high-speed internet access with a quiet work environment. To achieve the same flexibility as Uber, agents would have to train with dozens of different companies, which is not realistic.

Given the long-running prospect of COVID-19, contact centers face perhaps their biggest challenge yet to flatten the curve of disruption caused by spikes in demand or reductions in agent supply. The best ones will do so by doing these three things:

Deploying AI

Voice conversational AI—or a virtual contact center agent, delivering natural, human-like conversations on the phone—can be your first line of defense. Virtual agents are an efficient way to resolve repetitive calls, which are usually the exact types of calls that occur during call spikes. For example, a food delivery company may get hundreds of “Where is my order?” calls during bad weather. Or in today’s case, airline companies are getting bombarded by calls from customers asking to rebook or cancel their flights. These are precisely the types of calls that AI-powered virtual agents can answer with zero wait time regardless of call volume.

Elevating human agents

Removing mundane work from human agents will give them more energy and time to focus on more complex customer service issues, and to be more creative in doing so. Imagine you just dealt with 50 calls, each lasting less than 5 minutes, and provided similar solutions to all of them. How creative would you feel, when the complex 51st call comes in? Your brain will register it as “another one of those,” and take it on autopilot, resulting in an “another one of those” experiences for the customer as well. Using a virtual agent to take on your transactional calls will open up the opportunity for more higher-quality interaction between brands and consumers. ATMs did the same for banking. They now handle the most mundane tasks like check deposits and cash dispensing, which frees employees to provide higher-value services like financial planning and investment strategies. One unexpected impact of automation in contact centers will be less attrition among human agents because the work will be more satisfying, saving companies a fortune in hiring and training fees.

Enhancing service with greater insight

With conversational AI, contact centers will amass more data and insight into each customer interaction. Human agents are focused on the caller, and not on note taking, but virtual agents can do both at the same time. They’ll capture everything, consistently, adding up to call transcripts that can be easily searched and analyzed. This gives companies full visibility into why people call, how they need help, and what types of responses work best. Virtual agents are, in fact, uber-agents, becoming smarter as more data informs their ever changing responses. When conversations between agents of any kind and customers get better every day, customer service gets better, too.

Customer service is changing forever

When people think of the ways in which technology will change customer service, they often think of human agents being displaced by chatbots programmed to answer frequently asked questions. But with the rise and refinement of conversational AI, look for faster customer service, improved customer satisfaction, and more personalized customer interactions. Just like the ATMs didn’t replace your bankers, virtual agents will take on some of today’s tasks for human agents and free them up to perform more valuable tasks for their customers. 

How can Replicant.ai help?

Drop us a line at replicant.ai. We would love to show you how we are working with customers today to help them recover from the impacts of COVID-19.

The post 3 Ways Conversational AI Can Help Contact Centers Recover from COVID-19 appeared first on Replicant.

Most of us have learned about supply and demand in Economics 101. In a perfect world, increased demand leads to increased supply, and equilibrium is reached. Nevertheless, our world is far from being perfect, and friction comes into play. Think of the old days of ordering a taxi. Remember? You left a restaurant after a fun Saturday night, called a taxi and was promised it would be there in ten minutes. Twenty minutes later it started raining and your ten minutes became an hour. It is no one’s fault — it is a simple case of fixed supply (taxis) and elastic demand (Friday night + rain). We accepted taxis’ unpredictability as a given and just complained about it to our friends – that is until Uber came along and invented surge pricing.

Uber added important elements that changed the supply and demand curve. It added vehicles and drivers (more supply), but also created a mechanism to incentivize drivers to leave their homes and drive when they are most needed, by paying them more in times of high demand. Now, when demand increases, supply increases as well, until equilibrium is reached, and you can be picked up from this same restaurant in 5-10 minutes as expected.

So, why can’t your call centers be elastic? Why can’t you have trained agents that are called to work during peak hours? Flight cancellation due to an event like the coronavirus? No problem. Just pay 1.5X, and agents will be there; but, reality is far more complex than that.

Agents are not drivers. Driving a car is a generic skill that doesn’t require specialization. Google Maps gives drivers the ability to navigate any city regardless if they’ve been there or not. Drivers can easily move between cities, and even switch between driving tasks (i.e. delivering takeout, transporting people, or picking up packages) as demand changes. 

Agents on the other hand, are highly specialized and require company specific training. Agents need access to internal systems, constant updates on protocols and scripts, and high speed internet access with a quiet work environment. And, to achieve the same elasticity as Uber, agents would have to train with dozens of different companies, which is not realistic, of course.

So what can call centers do to achieve greater elasticity during times of uncertainty?

• Invest in workforce management software and best practices: Your first line of defense is implementing workforce management (WFM) software. WFM may not help with unplanned spikes but it can help create a solid baseline and prevent things like long wait times on a Monday morning when customers tend to call in on their way to work. If you are understaffed due to a lack of detailed planning, your ability to deal with spikes is limited at best which is why you need the right infrastructure in place.

• Have a work from home contingency plan: The coronavirus crisis is a perfect example; imagine you’re an airlines company – if there’s an increase in call volume from customers calling to cancel or reschedule their flights and your agents can’t come to work due to exposure risk, you now have more demand, less supply, and the result is tweets like this. Before asking people to work from home, you should 1) know who can work from home and 2) set agents up for success when they do work from home by ensuring access to high-speed internet and quality hardware and software.

• Develop a Maslow’s hierarchy of needs for customer service: Just like emergency rooms triage and prioritize patients, you should plan in advance for the types of calls that are most important to your business and customers. This means preparing agents to receive the highest priority calls, and deflecting the less urgent ones. Conversational AI (more on this below) can help to fully resolve a variety of transactional calls so that your agents can focus on the most important issues. 

• Use Artificial Intelligence (AI): Conversational AI on the phone (AKA Virtual Agents) is an efficient way to resolve repetitive calls which are usually the exact types of calls you get during call spikes. For example, a food delivery company may get hundreds of  “where is my order?” calls during bad weather. Or in today’s case, airline companies are getting bombarded by calls from customers asking to rebook or cancel their flights with the outbreak of Coronavirus. Similarly, a utility company may get calls about power outages following a major storm. These are precisely the types of calls that AI-powered Virtual Agents can answer with zero wait time regardless of call volume, allowing human agents to focus on more complicated cases. Conversational AI can act as your first line of defense, offering elasticity, resolving cases faster, and sending the most complex issues to your less overwhelmed customer service agents to increase customer satisfaction. 

• Prioritize self-service: Now is a great time to ask, “does this need to be a call?” In many cases, there is no symmetry between a revenue generating  transaction (i.e.ordering an item, booking a flight) and a cancellation (i.e. returning an item, canceling a reservation). For example, it took me 2 minutes to order a Nest Lock online, but to return it, I had to call customer service and spent 15 minutes on the phone with an agent. It takes 1 minute to set a password in a bank app, but it requires calling a bank to reset it. In both cases, a small investment in self-service like increasing in-app functionality can pay off handsomely and eliminate unneeded calls for your agents.   

There is no one strategy to deal with expected and unexpected call spikes, but deploying these strategies can help provide better customer service year round so that you can weather this storm and the ones to come. Have you implemented any of these strategies? Have you tried any others? Click “Let’s talk” on replicant.ai to drop me a line, I’d love to hear your thoughts.

The post Customer Service in the Time of Corona appeared first on Replicant.

Humans are naturally great conversationalists. After all, we have been conversing for 60,000+ years so it is second nature to us. We have spent the last 140 years mastering phone conversations and now understand the impact of basic phone etiquette like unnatural pauses, unplanned background noises, and human emotion to conduct effective calls. Machines, on the other hand, just got on the phone a few years ago so not much is programmed on how to make them great conversationalists.

At Replicant, we are teaching machines to solve tier 1 customer service issues on the phone, and we made it a priority to figure out what makes enjoyable machine to human conversations. Just like Asimov and his three laws of robotics, we created our three laws for human to machine communications and we cannot wait to share them with you.

Conversational Speed is Everything

When was the last time you failed to get a response for 7 to 10 seconds after asking a question because the other caller was on mute? Probably yesterday. There is a reason you remember cases like this; they feel strange. We expect people to respond instantly or we get frustrated. 

Now imagine a conversation where every time you finish talking, it takes 7 to 10 seconds to get a response. This is actually quite common with most machine to human conversations – machines “add” audio files of fictitious typing or paper shuffling to make pauses feel more natural because solving latency for machines is incredibly difficult. 

At Replicant, we understand that customers expect fluid conversations so we designed our technology to respond in under a second, just like a human would. This required building a state-of-the-art telephony system, an AI brain that can think in milliseconds, and many more subsystems that work in concert to ensure conversational speed.

Conversational Accuracy is Equally Important

Even well-trained agents do not always understand everything customers say. There may be background noise, language barriers, or everyday distractions during customer calls. Nevertheless, we expect the person on the other side to understand us and we become frustrated if they do not.

Low accuracy with an IVR or AI system also leads to frustration; customers try to “game” the system by guessing the correct key words like AGENT or RETURNS to speak with an agent faster. The moment this happens, the conversation ends, and the shouting match begins.

Yet, many AI systems are unable to have fluid conversations because their underlying models lack processing speed, accuracy, and most importantly, contextual awareness. In order to infer meaning in conversation, one must understand the full range of responses. This is very easy for humans as contextual awareness is gained through everyday conversational experiences. 

However, in the context of the “Thinking Machine”, it may ask, “Have you seen a doctor for this condition in the last twelve months?”. If the response is, “I went to the clinic last week”, a typical machine may have trouble understanding this as a nuanced yes. It is obvious to our human ears, but only a sophisticated machine would understand this too.  

Another important element is the ability to constantly retrain models as they learn to improve contextual awareness. Imagine if a caller is asked, “Are you ready?”, and they respond with, “Bring it on”. It is unlikely that a conversational model will have been trained on this reference so it must quickly ask a follow-on question to progress the conversation. Once the “Thinking Machine” gets a clear yes, this confirmation can be used to automatically retrain the model to recognize associated phrases so that it becomes smarter over time. Without a robust, continuous learning system, even a well-intentioned machine will quickly lose context.

Engineering an Expressive Voice is a Necessity

There is one remaining quality that we expect in conversations and that is to speak with someone that has an emotionally, in-tune, and expressive voice. It does not have to be a perfect voice with Hollywood-like quality, but it should be expressive. 

It is hard to stay engaged during long conversations when emphasis on key words is weak, questions are not always clear, and the voice is monotone. It can be overlooked during a short voicemail recording, but not for full length conversations. While what is said, and how fast it is said are far more important, having an expressive voice is the icing on the cake, especially if you wish to run a meaningful conversation between a machine and human.

Replicant’s speedy, smart, and engaging “Thinking Machine” can help you solve tier-one customer service issues on the phone in no time. We relentlessly focused on all three of these guiding principles when we built Replicant to create shorter, more effective calls to delight your customers. We hope that you too will be pleasantly surprised to see how much you enjoy speaking with Replicant. Visit our website to listen for yourself.  

The post What Makes a Great AI Conversation on the Phone? appeared first on Replicant.

At Replicant.ai we use AI to automate tier-1 customer service calls, and we create AI-powered virtual assistants for our customers. Because of that, we spend a lot of time studying the science of a great AI conversation, in addition to focusing on the dialogue and design principles of a profitable discussion. I wanted to share one of the frameworks we use to design effective and rewarding conversations so you can make sure that your customers are satisfied. 

So what makes a great conversation?

One of the most useful tools we’ve found to facilitate a good conversation is Grice’s maxims. Paul Grice, a British philosopher, came up with the following maxims for a successful conversation:

1. The maxim of quantity, where one tries to be as informative as one can possibly be by giving as much information as is needed to solve the problem, while excluding extra information.
2. The maxim of quality, where one tries to be honest by not providing information that is false or not supported by evidence.
3. The maxim of relation, where one tries to be relevant by saying things that are pertinent to the discussion.
4. The maxim of manner, when one tries to explain his/her thoughts in a clear, brief and orderly way, therefore avoiding obscurity and ambiguity.

Now, reflect on a great customer service call you’ve participated in. You can easily see how Grice’s maxims are relevant:

1. The maxim of quantity: Provide the customer with all the information they need, without burdening them with superfluous details.  “Your purchase is still under warranty and we will replace it at no charge” is an example. On the other hand, you would not want to say, ” You wouldn’t believe what I had to do before I got my manager to agree to replace your item for free. I had to email him 10 times.” Both responses contain the same truth and information, but adding unnecessary information may hurt your business. 
2. The maxim of quality: Honesty is key in customer service. There is no point in saying, “This has never happened before”, when people can google customer reviews or complaints and see that it has indeed happened. You may think the customer service conversation you have is contained in this call, but the caller is exposed to the experience of many others and can easily share the content of your conversation. Telling the truth requires less creativity, therefore creating a more consistent experience between agents. For example, saying that “Your shipment is delayed due to an error in label printing. We are sorry and will give you a $10 credit for your next shipment” vs. “Your shipment is delayed since you provided the wrong address.” Altering the truth to avoid angering the customer rarely works and often creates a negative outcome. 
3. The maxim of relation: the more you understand the callers and their needs, the more you can frame the conversation to make it relevant and relatable. If your customers are detailed-oriented engineers (or event planners), you may want to provide detailed explanations. If you sell services to busy executives, you may want to keep the conversation short and to the point. In any case, do not share random information that is irrelevant to the conversation, unless it is well crafted and serves a purpose (e.g. building a relationship)
4. The maxim of manner: Oh clarity! This is where conversations can go really wrong. You know things that your customers may not know so make sure you walk the customer through the explanation with a logical flow. Sharing your knowledge with customers makes the difference between a short and effective call or a long and confusing call. You also want to make sure that your call script and agent training correspond to the audience: Are you contacting PhDs? If so, devise a more sophisticated conversation. However, if your product caters to a very wide audience, you may want to aim for clarity by using the USA Today’s level of English. 

When Replicant.ai constructs call flows for our “Thinking Machine” we take Grice’s maxims into account, making sure we understand the audience, the customer’s case, any problems that could have initiated the call and the best way to solve the problem while also making sure that customers get a consistent, honest and clear response every single time. The same best practices we use for crafting great AI conversations can be used to craft human to human conversations. After all, Paul Grice was born in 1913 and probably didn’t envision that his principles would be used a century later for AI virtual agent conversations. If you incorporate the Grice’s maxims in every conversation, you will discover that the metrics you are so worried about with your call center, take care of themself.

The post What Makes a Great Conversation? appeared first on Replicant.

Implementing these new technologies into the call center allows businesses to rethink their customer service delivery, making sure that they are providing optimal service while utilizing their resources—including their live agents—more efficiently. But when incorporating automation into your center, what should be automated?

When deciding how to merge automated virtual agents into your call center, you should consider this integration from three angles:  the customer, the live agent, and the company. Let’s explore these key areas further.

How to Optimize Your Call Center with Automated Virtual Agents

When revamping your call center with automated virtual agents, you should consider which calls should be routed to virtual agents and which should be accelerated to human agents. In making this determination, you should consider the complexity, the urgency, and the emotionality of the customer’s request.

The more complex, urgent, and emotional a customer’s request, the more a human agent is required for hands-on creative problem-solving. However, for more routine inquiries, such as a change of address, bill payment, or balance inquiries, virtual agents are the perfect solution.

For example, if a customer calls your service center with food poisoning or a travel emergency, a virtual agent will frustrate the situation, as it cannot empathize with the customer or provide unique, creative solutions to the problem. Upon receiving a more sophisticated customer service call, the bot can detect the customer’s urgency through technologies such as artificial intelligence and natural speech recognition, quickly routing the call to a live agent.

For more routine and repetitive calls, the ones not requiring human creative problem solving, such as claims management, collections, and FAQs, route these calls to your automated virtual agents. Virtual agents can handle these calls faster than a human, while making fewer mistakes. Further, the virtual agent will not become frustrated with such routine, monotonous tasks. The virtual agent’s consistency in handling these requests will not only create more positive experiences for customers, but it will also produce shorter wait times and more resolved issues.

How to Enhance Your Agents’ Jobs

According to Gallup, only 34% of U.S workers are engaged in their jobs, meaning they are enthusiastic and committed to their work and employer. Although this is the highest engagement percentage since Gallup started reporting on this in 2000, that leaves 66% of the workforce unengaged, with 13% representing those with “miserable” work experiences. Some issues causing lack of engagement include poor relationships with managers, sub-standard benefits packages, and the absence of employee coaching. Additionally, employees who are tasked with monotonous, repetitive jobs, where they don’t have the opportunity to use their skills and strengths, also contribute to job dissatisfaction.

In call centers, repetitive, rote calls can cause live agents’ frustration and discord. Answering questions like a robot does not add to most people’s days. These are empty-calorie calls, leaving live agents with no leeway to use their talents and creativity in difficult problem-solving situations. Leave the robot questions to the machines.

By directing these routine calls to automated virtual agents, live agents can focus on empathetic, creative, high-level problem-solving. In a recent study, 100% of respondents stated that customer experience cannot be achieved without engaged employees. Live agents will feel a sense of well-being, knowing that their contributions to the customer experience are valuable, giving them a sense of ownership in their jobs.

Additionally, engaged employees contribute to the company’s bottom line. While boosting your company’s customer experience, you’re increasing your company’s profitability. With engaged live agents, Gallup found that companies experience 21% higher profitability, better retention, and better customer engagement.

How Your Company Benefits from an Optimized Call Center

With call centers serving as the primary means of communication between a customer and the business, companies need to enhance their customer service offerings continually. Today, customer service is experiencing a digital transformation, from merely inputting data into a computer to full customer experience management, with the ability to coordinate and personalize the end-to-end customer experience, at any time and on any channel.

In 2019, 80% of organizations plan to compete on customer experience. A frustrating, slow, or inattentive customer-service frontline detracts from the customer’s overall experience, causing more harm than good.

Give yourself the edge on industry competition by focusing on your call center. Companies that are customer experience-led see 1.9x return on spend, 1.9x higher average order value, 1.7x higher customer retention, 1.6 higher brand awareness, and 1.5x higher employee satisfaction.

With a customer experience-led call center incorporating automation, you’ll see higher consistency, better accuracy, more compliance, and faster resolution times. Virtual agents are better and quicker at sorting through data while recording it. Virtual agents are great with a script. Virtual agents will get a change of address, billing address, or shipping address right every time, by instantaneously running customer addresses through geo-databases for confirmation. Virtual agents come loaded with data and information and don’t have to open multiple computer windows or place the customer on hold to review and find relevant facts. And, virtual agents do not get offended or frustrated. 

Machines cannot do everything humans can do, but they can help companies listen carefully, think quickly, take actions, and reply fast enough to be conversational and effective. Virtual agents can instantly process data while solving problems and reduce cost (since computers don’t need to be hired or on-boarded and never call in sick), while offering high-touch customer service.

Virtual agents working alongside human agents can help you scale your business, remain open 24/7/365, can solve problems more consistently and can make your customer’s efforts effortless. With the help of your tenacious human agent problem-solvers, you can enhance and optimize your call center. While your virtual agent handles routine calls quickly and effectively, your live agents are free to handle more complex customer roadblocks while approaching their jobs with more flexibility and creativity.

By training human employees to handle high-level customer service while having computers handle more routine (and let’s face it, boring) issues, companies can invest in the local workforce while providing a customer experience that keeps consumers coming back for more. If you want to find out more about how to strengthen and add value to your customer call center, give us a call.

The post Call Center Automation: What Calls Should You Automate? appeared first on Replicant.