Who among us hasn’t called a contact center at some point for customer service, tech support, or some other matter? Usually those interactions are mundane, unremarkable, and ultimately forgettable. You get what you need and go on with your day, blissfully unaware of what transpired behind the scenes to provide you that friction-free encounter. It turns out however that automation & AI often play a key role in delivering those satisfying outcomes, and NICE Ltd. (formerly Neptune Intelligence Computer Engineering) is a key player enabling those experiences.
Oded Karev runs NICE’s RPA division, and has a lot to say about how the attended & unattended real-time automation they’ve innovated has digitally transformed their contact center customers. He joins us to share his insights about some interesting use cases, whether or not all the various forms of automation will eventually merge into one singular platform, and briefly talks about NICE’s next major innovation – robots building robots.
Guy Nadivi: Welcome everyone. My name is Guy Nadivi and I’m the host of Intelligent Automation Radio. Our guest on today’s episode is Oded Karev, General Manager of Robotics Process Automation at NICE Limited. NICE, which I recently learned was founded as Neptune Intelligence Computer Engineering, hence the acronym NICE, is one of the world’s biggest providers of attended and unattended automation. NICE counts 85 of the Fortune 100 as their clients, who use their automation platform for everything from streamlining call center workflows to combating money laundering and terrorist financing. With that range of use cases, we thought Oded would have some very interesting perspectives on the automation space, so we’ve invited him onto the podcast today to share his insights with us. Oded, welcome to Intelligent Automation Radio.
Oded Karev: Thanks a lot for having me Guy. Great to be here!
Guy Nadivi: Oded please tell us a bit about how you got into the field of automation and the path you took to become CEO of NICE RPA.
Oded Karev: Well, I joined NICE back in 2012, after a long career as a strategy consultant. I spent a long tenure with Accenture Strategy Consulting, crossed the line to the vendor side and helped NICE shape the corporate strategy for a few years. Around 2016 at the corporate strategy team, we identified that the RPA market is booming and starting to grow and NICE have both the technology and installed base to be a significant player in this market.
And therefore there was a decision to create a dedicated line of business, taking that technology that NICE acquired from eglue back in 2010 and making it a dedicated business. There was an opportunity for me from the NICE CEO from Barak that came to me and said, “How about you do another move from being an internal director of strategy to actually leading a business.” And I had the opportunity to lead the NICE RPA business since 2015/16, so five years now. Probably the most exciting times in the RPA market, with all the growth and the action and the money flowing. And evolution of the technology going mainstream. So both personally and from a company standpoint, it’s been great.
Guy Nadivi: What are the nuances people should know about, when considering attended versus unattended automation?
Oded Karev: I’m not surprised you direct the first question to me about attended. NICE is known as the leader in attended automation even more so than unattended. This is where we came from and we started. And there are a lot more than even nuances, right? Attended automation is so much more complicated than unattended. And the gap, the technology gap and the technology gap leads to the art of the possible and how many use cases you actually can benefit from is relatively big. I recently saw one of my competitors referring to attended automation as, “Automation triggered by the employee from the desktop.” Which is true, but it’s like describing a car as something with wheels, okay? It also have an engine, it also have a lot of other stuff. And triggering the automation from the desktop is probably a small and not the most important capability of attended. And the bigger differences of attended automation versus unattended: It’s not just the control of the employee triggering it, it’s actually the ability of the robot to react, to trigger itself.
Using advanced AI and image processing and so on, our robots monitor the screen in real time. And our robots decide when they need to start to automate. Now, this is a huge difference between waiting for the employee to press a button to launch an automation, not only because you’re dependent on the employee when it comes to change management and seeing value and making sure he adopt. But there are a lot of use cases you don’t want to leave it to the discretion of the employee. For example, any compliance-related process. If you want to ensure compliance adherence, if you want to ensure a certain process step happens, if you want to ensure a certain disclaimer is being read. You want the robot to identify which process is being executed and automatically either execute the process or provide the guidance.
Same goes for if you want to turn service into sale, what we call service to sale, or the next best action mystery. You need a robot to identify the customer, to identify the process, to identify the service interaction, and to come with the right next best action proactively. Not to wait for the employee to ask for that. And it also goes for accelerating and reducing handling time. You want the robots to react to what the employee is doing. And this is true men working with machine. How to do that, how to do that without slowing down the performance of the machine and the computing power? How to do that without people getting annoyed that the robot is taking over? How to get to this harmony of men working with machine? Well, you need about 20 years of experience with that, we have 19. So we’re close.
Guy Nadivi: Oded, NICE’s background is in contact centers, which I believe is unique among RPA companies. What advantages has that provided you, vis-a-vis other RPA tools?
Oded Karev: It’s a two edge sword, right? Definitely the very complicated use cases coming from the contact center. Our automation technology started from automating processes in the contact center, working with agents. The real time aspect of the automation, it cannot be asynchronous [sic]. It cannot be sent to a server somewhere in the back end of the IT basement that anything would happen on the server. It needs to happen in real time because the contact center processes are usually real time. Once you call the contact center, the process happens here and now, there’s no room for error. And the aspect of the human or the involvement of a human is mandatory, right? You call the contact center, you speak to a human. There is no way now to offload it to an unattended robot. You must have the ability to automate real time on the screen, man with machine.
So it provided us a lot of value, the domain expertise. The access to the customer. The NICE install base. The understanding of the nuances of the contact center. The ability to automate in real time. The technology capability of everything I described before. We even used a robot for performance analytics for contact center agents. Because we’re installed on the contact center agent screen. Because we monitor the screen in real time, we said: “Why don’t we aggregate this data and run analytics on it? And we know what the contact center agents should and shouldn’t be doing. What is productive, what is not productive. And it can provide statistic on it and measure them on it.” So it’s been truly great for us going with that to the contact center. Naturally, our reach to the corporate shared services: HR, IT, finance or even our reach to the back office have been somewhat harder, because most of our domain expertise and install base and go to market, is focused on the contact center. And though we believe that our technology is the best platform out there to automate end to end, we have seen greater success in contact centers. Guy Nadivi: How has the pandemic affected NICE’s business?
Oded Karev: Well, NICE overall have seen a great growth with the pandemic, right? NICE is the global leader in a contact center in the cloud. Setting up a contact center of thousands of agencies in a week, or even less than that, 48 hours is something that only NICE have the capacity and the capability to do. So all over the world contact centers were reason to support the vaccination effort or to support the question employee people… citizens have for the pandemic and so on. So NICE have grown very well. Even within my small four corners of the automation, we have seen a lot of interesting use cases and a lot of interesting growth. Starting with helping with the National Healthcare society in the UK, where they needed to automate to allow the extra capacity for nurses to provide care over the phone. All the way to support work from home. As I explained my software, the attended robot, they work distributed. They work completely distributed on many, many desktops. And when it comes to scale my biggest customer had 40,000 desktop. After that, I have 32,000 desktop robots. So truly large scale customers. All of them are very much distributed. When you think of work from home, it becomes very important. How do you monitor the screen of the employee that is not in the office anymore? You don’t have the team leads to sit behind his shoulder and see what he’s doing. It’s not only governance and monitoring. How do you help him? Usually in the contact center, we have the swivel chair effect. The agent turns around to the agent next to him and ask him, “How do I do this operation? I have a customer is asking me about something I never did before. Which button to press, how do I progress? What am I allowed not allowed to do?” When you swivel the chair in your home office, you see a wall. You see your wife. You don’t get anyone to ask. So the ability to provide guidance via attended automation, and having the robot as a personal assistant to each and every employee is a platform to ask questions. To get guidance. To listen to management updates and so on and so forth. So the work from home increased the need of agent guidance and we enjoyed it very much.
Guy Nadivi: You mentioned working from home and the pandemic has obviously disrupted organizations in profound ways. Forcing many to work from home instead of an office as just one example. Do you think the pandemic has made resiliency as important or perhaps even more important than things like cost savings as a justification to invest more in automation?
Oded Karev: I do. I do. I think that when the RPA market started, it was all about AHD. Like reducing handling time, efficiency, efficiency, efficiency. Now you understand that you need extra capacity. And it’s much easier to turn up a few more robots versus recruit and hire and train more employees. You also understand that employees are more flexible. They can react better to changes versus a robot that is more rigid in how we execute stuff. And therefore you need your human workforce to do stuff, to be more agile versus your digital workforce to continue and execute the repetitive and mundane processes. We have seen adoption for increasing capacity just not for saving, but just to keep the lights on. Having the business operated. We have seen companies who were really impacted by the pandemics. I have few customers now in India, big BPOs that they unfortunately suffering a terrible second wave and employees cannot come to work. Either they’re sick themselves, or they’re taking care of a family member and immediately your workforce goes down. The ability to enjoy or to augment digital workforce and physical workforce have never been more important.
Guy Nadivi: Oded, Can you speak about some of the more interesting use cases NICE RPA has been applied to, and the results you achieved?
Oded Karev: For me, everything is interesting. I’m extremely excited about hearing new use cases and hearing how creative our customers are getting in leveraging our platform. And every customer is different, right? We provide an automation platform and a robotic platform and every customer ask the robot to do different stuff. Naturally, the more things that they are done with the platform that are aimed at the benefit of the world and at the benefit of people are better. I can give you an example of a very large tax authority from a European country, that took a decision to improve service level of the citizens of this country. Now, I think that’s revolutionary thought. I would never see my own country or I don’t imagine my own country deciding to improve service when it comes to tax collection. But they decided to be more citizen-centric. And we were able to automate processes, taking them from a three minutes to 15 seconds – customer queries. So you would call, you would ask the text agent a question: the government agent. He would put you on hold for three minutes he would need to aggregate data. We were able to automate that, to give him the relevant data in 15 seconds. Reducing the handling time. Improving the agent experience. Improving the employee experience. And I found it great. Just going back to what we said, they’re a government agency. They’re highly unionized. They’re not looking at efficiency in the traditional way of cutting costs. They looked at the efficiency of better experience to the employee, to the citizen of better experience for you not to wait long time on the line because the calls are shorter. It’s truly creative. And of course, everything that is done with AI and today, amazing things are done with AI. For the same government agency, we recently were able to automate the handwritten analysis. So someone is sending a form in handwriting and our partner in OCR engine takes that and read handwriting and convert it into a digital data that our robot can automate on. And the accuracy level of that conversion was higher than a human typing. So in the past, they would give… There was a big administrative team receiving the forms in handwriting and typing them in. So the data would be digital and you can automate it. Now we’re able to convert it. Not only without people just doing mundane task of typing, but also with a better accuracy and lower error rate. So AI is becoming great. And we recently launched our vision of robots, building robots. Which I think is the most exciting thing we did from an innovation standpoint in the RPA market. So we started by building a new design environment and how you build robots in a very easy way. We continued with the discovery tool, which is complete unsupervised machine learning, truly auto discovery, where we deploy our robots. The robots run for 10 days, map all the processes and provide back a dashboard with recommendation of what to automate. And now we just released a very magical button where you press on it. And the data from this discovery tool is converted to our design environments, to the automation building tool. And you literally have the robot identifying the opportunity and building the basic or the skeleton of the automation. That then goes to testing and exceptions and so on. But the vision of robots building robots is very exciting.
Guy Nadivi: Interesting. In an article earlier this year, you stated that, “Our most scaled and successful customers have established COEs that cover the full automation life cycle.” Like NICE, just about every automation vendor encourages their customers to create an automation Center of Excellence, as part of their digital transformation initiative. However, EMA research reported that only 7% have done so, despite there being persuasive evidence that COEs play a major role in an organization’s automation success. Why do you think COEs haven’t proliferated as widely as they should?
Oded Karev: I think it’s a matter of market maturity, all right? Originally the promise of RPA was almost too good to be true. And some vendors… And I say that with a hint of criticism. At NICE we avoided that because we are publicly traded company and also we are a portfolio company. So I cannot over promise and I cannot sell dreams in my line of business, because then it would hit my colleagues when they’re trying to sell their products the portfolio company. So I was more regulated internally, but some of the players in the markets over promised to the capability of RPA. And oversold the dream of how easy it is and how simple it is, and how with a minimal investment you drag and drop and immediately you have an automation. And then customers face the reality. You’re an enterprise software organization. You’re a large enterprise. You have a big IT. You have cumbersome processes. This is not magic. Like any software, like any transformation you need to invest in it. And the best way to do it, you can start with a partner, you can start with an SI. But either you take this SI in a long-term engagement as a managed services, or you take ownership over the software. You take ownership over the solution. And that is what we have been recommending to our customers from day one. And as you said, a lot of the other RPA vendors understood that long-term, in order to scale you need to explain to the customer that it’s not one and done. It’s not, you install the automation and everything works. You need to continuously add automations. Manage the automation. Maintain the automation. Advocate the ROI. Build your automation pipeline and so on. And as the market mature, more and more vendors are preaching that you must as an enterprise take ownership of your automation program. As more vendors preach to that, then eventually the adoption would grow. And I believe that all scaled RPA organizations would have a COE, either their own or outsourced to a partner. But there’s no other way to be successful.
Guy Nadivi: NICE specializes in RPA, which is one type of automation, but there are many others including: DevOps, IT process automation, Runbook automation, Security Orchestration Automation and Response, also known as SOAR, Workload automation and others. There’s a belief among some that the industry is heading towards a convergence point. What Gartner has called Hyper Automation and Phil Fersht of Horses for Sources has termed Integrated automation. I also had a recent guest who said that ultimately integrations combined with AI will mitigate the need for automation all together. Where do you think things are heading for the automation industry?
Oded Karev: I think that putting all these technologies in the same bucket is a little bit misleading, because RPA is one approach for automation, right? It’s a very different approach than the IT orchestration or testing automation or BPM or SOAR or other stuff, right? It comes in and says, “A human would mimic or a robot would mimic how a human would execute the process.” And I’m not changing anything. And I’m teaching the robot with a very simple layer. And most of RPA automations are for limited time, two, three, four, five years. Eventually, as IT matures, as another system is being replaced, you’ll replace your ERP, you’ll upgrade your CRM, you’ll add another system, you automate more processes. And the need for these automation decreases. In a lot of cases the need for another automation rises and you use the robot for something else. But robotics is usually not a long-term solution to replace a SOAR architecture or IT architecture. You use it as a tool to provide fast ROI, augment the workforce, kind of a quick and dirty solution. Where other solutions for automation are more salient and may be for long-term. So I doubt we would see a convergence, also because of the inherent difference in the technology. And how you maintain the technology. And how you approach the technology. We may see a convergence with what we call the low-code domain. Where you have platforms that don’t require a lot of effort to automate from different standpoints and RPA is just one of them. And there are a lot of other capabilities. Maybe there’d be a convergence there, but I don’t see a full convergence of the entire automation world. Let’s remember any software out there is automating something. The calculator is automating how we do math, right? In a very simple way. And the CRM is automating the workflow of trouble ticketing and services processes. Eventually every software out there is automating something. And the business process platform take very different approaches and require very different technologies. So I doubt we would see such a convergence because the inherent difference in technologies. I do think… I don’t know who said it. One of your previous guests, you said: “Around AI and integration would solve everything.” It’s interesting, kf we’ll be able to get to that point, just adding layer over layer, over layer, instead of replacing the underlining technology. I think it would end up creating something that is unmanageable, but maybe it will be fixed. Time will tell.
Guy Nadivi: Oded, for the CIO/CTOs and other IT executives listening in, what is the one big must have piece of advice you’d like them to take away from our discussion, with regards to automation at their organizations?
Oded Karev: One thing? I think we just touched it. I had a lot of conversations with large Fortune 100 CIO/CTOs that explained to me that the technology is dirty. That the technology is not clean when it comes to enterprise architecture. And I agree. But to do now with full transformation, all the way down to the core systems, and replacing CRM and ERP and billing and all the landscape, it’s a multi-billion project. Multi-year project that you know how you start and you never know how you finish. And RPA is almost a way to fund that journey. Because it’s quicker. It brings ROI. You can use it for a certain amount of time. You can leverage the license over and over again. So it’s definitely something you need to consider. And when you think of how complex our world is becoming and going back to the contact center, especially for customer service representative. What we have seen recently is that a huge adoption of digital, right? Especially in COVID people go online and buy more. And people schedule and talk to the doctor via the Zoom Meeting and so on and so forth. But when they need a human, they no longer go to the branch. They reach out, they call the contact center and do a video call, they chat, they need a human. So the calls become more complicated. And actually counter-intuitively, the contact center role is becoming more important. As the physical high stores and branches and service centers importance is going down. And the complexity of the interaction is going up. You must provide tools that would help your agents. Customer experience is everything today and relying on old systems or saying, “I’m going to transform my entire digital service.” And hope it’s going to take a year, it’s not. And only by using attended automation and providing each and every CSR with his own virtual assistant, they have the right guidance in real time to provide this excellent service you expected.
Guy Nadivi: All right. Looks like that’s all the time we have for on this episode of Intelligent Automation Radio. Oded, one thing that’s been intriguing to me about NICE is how your automation leadership emerged largely out of a focus on call centers. I think most people view call centers as playing a mundane role in the greater customer service function. Yet it turns out that the work which goes on in call centers, as you’ve pointed out, is ripe for the kind of automation innovation which NICE developed. And is now extending to other areas of an enterprise. We look forward to seeing your continuing advances in the automation space. Thanks for coming onto the show.
Oded Karev: Thanks for having me. It’s been great, Guy.
Guy Nadivi: Oded Karev, General Manager of Robotics Process Automation at NICE Limited. Thank you for listening everyone. And remember, don’t hesitate, automate.