Driving public sector efficiency: How AI can help you better service your community

Good afternoon. Thank you, everybody, for joining our, webinar. Again, this is a, webinar number two in a three series webinar set with, Verint and TTEC. Hopefully, you joined the first one, which, spoke about FedRAMP and StateRAMP and, the wonderful world of that being pulled into, obviously, public sector. This particular webinar, we're gonna be speaking about, driving public sector efficiency and how AI and CX automation can help better serve your community without increasing resources. So, again, thank you for joining. I'll start off with my introduction of I am Scott Montgomery, and I am the vice president of public sector sled within Verint. I've been working, in this industry for over twenty five years, focused both on public sector and commercial, and, obviously, focused right now in mainly North America business. But, I'm also, having this webinar with Caleb from TTEC, and I'll let Caleb introduce himself. Yeah. Perfect. Thanks, Scott. And thanks again everybody for joining. I'm Caleb Johnson, vice president with TTEC Digital. I work on go to market and work very closely with many of our partners like Verint, like Google, like Cisco. So we're really excited to jump in these topics, talk talking about automation and AI. So, Scott, I'll hand it to you. And to kinda kick things off, we can get going. Thanks so much, Caleb. We're gonna have some fun today, and hopefully, everybody's gonna enjoy it. So the first thing I always love kicking off with these webinars is to kind of throw out our problem or at least throw out what we usually talk about a main problem is. And in this case, what we're gonna start off with is, the wonderful world of not enough resources, not enough to not enough to control with what we need to. Verint has typically deemed this as, what we brand the engagement capacity gap. And, really, what's happened here, it started, obviously, several years ago, and where we've seen it kick off ramp up is with originally with COVID. Right? And what we started to see is a higher increased demand with our citizens, and that's their expectation of how they, communicate and work with any agency. That's via, you know, voice, email, etcetera. But that demand has gone up significantly. They're expecting now all of these, organizations and agencies to operate very similar to what a commercial or an enterprise customer would. And, obviously, what we have a challenge with is we don't have the budget like we used to. Even if we do have a good budget, it's very hard to find the resources we're looking for. So no matter how we look at it, we have this gap of what the demand is from our citizens and what we're able to actually use with physical resources, human people resources. And the only way to really address that capacity gap is through AI, through technology. And that's where we start this off with, is talking about how do we bring that and close that gap. So I'm gonna hand it over to Caleb to continue the conversation about that. Yeah. That that's perfect. So one of the ways we look at this and and really to Scott's point on it's a challenge with resources, we start to look at also kind of operational strategies of we have all these new technologies. Where can we go apply them? And it's not always simply simple as just turning something on and it's just gonna work perfectly. There has to be an approach shift. And so part of what we've tried to help a lot of our customers with is this shift moving away from a reactive model where there's an event, someone's reaching out, they need assistance, and really trying to get ahead of that and move them up this maturity, curve, if you will, getting into more proactive service, and really all the way with this kind of utopian goal of preemptive service. But using data, using AI, using automation to be able to start to get ahead of our citizen events or moments of need and being able to use that tech to be able to reach out proactively, use the data that we've got to be able to send out messaging at the right time, leveraging self-service strategies so that, really, we don't ever need to get to a resource to actually get assistance. So not only does this really help with that resource gap that Scott's talking about, but it also at the same time, we're able to start looking at, can we actually improve the overall experience, improve that, that citizen satisfaction score, improve the overall, just journey that someone's going through, but also remove the friction from that experience. And so as we kinda get into some of these other topics, we'll kind of lean on that theme of where we can apply AI automation into these strategies to start shifting away from reactive and more into this proactive and kind of preemptive model as we support these different experiences. Absolutely. Cale, that's great. So what we start off with is, as we just talked about, where do we interject that AI? Where do we interject those pieces? And one of the basic places you have to start with is understanding your data. Right? We go back, and I mentioned COVID earlier, and we think about, you know, where did we originally interact with our citizens. And several years ago, right, you started thinking about majority of it, right, from an agency point of view was either voice, right, or maybe even coming into the actual, center itself. Right? So if we're thinking about, you know, a DMV or somewhere else, right, physically coming in. And there was a lot of just voice interaction and voice interaction alone. But, obviously, now, fast forward to where we are today, and everybody wants to communicate on multiple different levels. We're looking at not only communicating from a point of view of voice still. Right? But we see a huge digital shift, and we understand why. And that is, you know, now you're in email. You're in self serve. You have your chatbots, right, you're that you're interacting with. You have social media, right, that they're interacting through. And then you have all of, which we all are very familiar with, all of this data on the back end, right, between different interactions of, CAD platforms or CRM platforms or platforms that's hosting other data in regards to, you know, where you're sitting at, you know, your location, your address, etcetera. And all of this data matters because this data is what's gonna help drive or you to better understand where you can apply AI to help make that move forward and help reduce that, that gap as we said earlier. So now that we talk about kind of where that data is, we talk about why is that important to bring it all together. The first thing we kinda talk about is all of this data tends to be unstructured. Right? You know? And what we wanna do is we wanna create a structure around it, and we want a better understanding of that structure. We pull that into a, you know, weak or coin as our engagement management data hub. And that allows us to bring all of this stuff together and allows us to start normalizing it. And, again, why is that valuable? Because what you're doing is you're you're looking at how you're interacting with that citizen in a holistic standpoint, not just from a one individual voice interaction, but that same citizen can be originally starting with a communication with you via self serve or via a chatbot, move into maybe an email for a follow-up, tie in at the very end if they're unable to resolve what they're looking for as a physical phone call in. And all of that data matters for how we're looking at where we could apply AI, where can we reduce things, again, not just to help the citizen, but also to help the agent. Yeah. And, Scott, I'll just add on one thing to this because I think it's a a really great call out. When we engage with customers a lot of time, there's there's typically a tactical kind of need. Somebody needs to be able to go in, and we need to reduce volume into a contact center, or we need to solve this surge that was kind of unexpected. But in in a perfect scenario, kind of our guidance back to everybody is you really want to start at the data to be able to make decisions. So if you're gonna go in and kinda change some of this operating model or how you're really gonna interact with citizens, It's not so much about just throwing out a, a point solution solved for one use case, but it's really looking at, at this this large dataset that's now available and be able to make data driven decisions and let that guide a you know, sometimes it's a multiyear road map of automation, being able to set a foundation and build on that over time. So, I think you're spot on with this, this whole concept where it it has to be a data driven at the you know, approach at the very beginning as you start thinking about where you're gonna go apply AI automation in into that whole journey. So it's great. Absolutely. And and this is a great time to start bringing up examples. Right? So, what works best anytime I, we talk about this or or or educate anybody is is to give real live examples. And I'll give you one here where, it always comes up. A great example of why you look at multiple datasets. I was with a particular customer. We were having a conversation about how they were looking to implement some AI. And, really, what they were doing is they were looking to try to resolve one of their top business challenges. And, unfortunately, what they were doing is they were looking at just one set of data. They were looking at, in this case, how the interaction was from all of their voice interactions, particularly with a, with the citizens. And and they were they were basically mapping out their top ten next things they need to do. But what they weren't doing is actually looking at the rest of the dataset, and the rest of the dataset we're talking about is the actual interactions they were having with those citizens from a chat as well as a email as well as a, self serve perspective. And, in the conversation, you know, we kinda challenged and pushed back and have them go analyze that. And what they found out is what they're originally gonna go after, right, on the voice side was gonna produce basically an x return in value because it was gonna resolve some serious business challenges they were having. But when they actually analyzed the rest of the data, they uncovered there was two or three other challenges they were having inside their other dataset that was far more valuable in a return on investment and created them far more citizen satisfaction by tackling those first. And, again, it just comes back to the point of, you know, when you're looking at this, data driven is really where you wanna go because that helps you understand the big picture, also helps you make your decisions in a in a more effective way. Yeah. For sure. So next is what we're talking about that is is why do we have an open platform where it sits at? And, again, why I talk about this or bring this up is to go back to the basics what we just said. Hey. We wanna interject AI. We wanna interject process improvement or automation. Where do we gotta start? First, we gotta pull that data in, which you see it sits here at the center of that engagement data hub. And what you see also sits at the center, and what you wanna drive is now the AI and automation. Because now now you have that data in there. We just talked about how you're gonna review that data. You're gonna understand where the potential of those business challenges are. And now you wanna take that AI and apply it into that business to help resolve or augment, you know, some of your workforce. And that's obviously the goal we have here. Yeah. So let's jump into some some use cases. We talked kind of around structure and kind of platform and definitely the the need for data, which I completely agree with. But here's some real world use cases where we see things going into application. I mean, I think for the last several years, there's been so much buzz around around AI and where you can go apply this, whether it's on the front end of that, you know, the citizen journey where they have that first moment of of reaching out to get support, or you start getting into kind of this, what we'll call the back end, looking at automation and ways to make things more efficient. And a lot of times, that's bringing data forward or simply making it more integrated. Right? But some of the easy ways to kinda remove conflict from the from the experience is around things like conversational orchestration. Just being able to give people kind of preference of channel where, they, you know, they call in. There's a voice interaction. You know, there may not be enough people to support all of those interactions. Being able to move that into other channels, whether it's, you know, SMS or it's chat or even into a social channel. And and, likewise, being able to take somebody from a digital channel and move them into voice when it really, really matters, but being able to leverage the right tech to be able to simplify that. And then also being able to tap into that, personalization, right, that into the individualized development concept where you have a lot of this data that's available. Sometimes it's sitting in silos, but being able to use that to route just somebody to the right person, that can be a huge win and really help reduce some of the cost as well as being able to use that same, information, you know, kind of that personalization concept to be able to say, alright. We know that there's gonna be somebody needs to refile something at the end of this month. Proactively using that to go out, send a message so that somebody's not late. Right? These are all kind of things that reduce inbound volume from the contact centers and be able to also just improve, kind of overall satisfaction because you're actually helping people out, giving them the information they need. We've also started to see AI really come in from, you know, this outbound concept, but being able to leverage all that data that you start to collect, from understanding the the cycles and surges of events and then be able to tap into this, all the citizen data to be able to start all this outbound communication and then opening it up so it's a two way conversation. Because there are times, I think, everybody's probably gone through this before you get a, you know, a text message reminder about something, but you can't reply back to it. Right? It's kind of a dead end. You need to call in if you actually wanna get support. So start opening up some of those channels so that we can put a virtual agent in those digital channels. You start to use a a lower cost to serve channel to be able to interact and enable self-service, you know, based on these proactive outbound events. That's a great way we started to see the the AI come into this. And then lastly, from a a security and a fraud perspective, just given all this information that we've been talking about, the more companies and, you know, agencies get organized with being able to integrate data into a way that you can actually tap into it to make decisions, you're able to start to kind of elevate the ability to understand who is interacting with you and then be able to start using that from an authentication standpoint, from a fraud standpoint, and really make sure that who the person is that you're interacting with is who they say they are. And so there's some great solutions that we start to see just massive advancements in, leveraging AI, leveraging even large language models on the back end to be able to have that conversation on the front, use it to be able to, use the data to be able to authenticate the right people, whether it's, you know, going down the biometric path or multifactor authentication. There's even some new things coming into play on that front. But all of that's really starting to improve this model and and and drive down the cost to serve in the contact center, which kinda goes back to Scott's original point about it's a this is all a race for resources. Right? It's about being able to have enough people to actually, handle the volume that's sitting out there today. So just some real world use cases we can continue to build on here over the next few minutes. Absolutely. Those are great examples. And, take the next step forward. One of the things I love to talk about you know, we mentioned AI. We talked about AI. You know, AI has been around for such a long time. You know, we've seen it all over the place. Right? We see it originally with chatbots. We've seen it, you know, both used in both positive and, you know, not necessarily, negative ways. Right? But the whole concept of of AI, it's it's kinda like that generalized word. You know, what we wanna focus on or what we wanna talk about is when we actually mentioned AI. Right? Really, one of the things we we think about is is as a bot per se, a specific bot. And a bot is not just I'm not just referring to a chatbot a chatbot. Right? I'm referring to, you know, what a bot is is something that's doing a single human function. Right? And it does one thing and one thing very well. And and why does that matter? Because, again, we start talking about AI. It goes back to the data. It goes back to several examples Cabel, Cabel just mentioned. Right? You're looking to help augment your workforce. Right? You're not looking necessarily to replace the workforce. Right? That's not the goal here. The goal is we know we don't have enough resources. So where can we support, you know, the citizens? Where can we support the actual agents, right, to help them in their day to day? Right? And every way we help them, right, it saves time, saves money, provides additional bandwidth for them to accomplish other goals, right, which is working with other citizens, etcetera. So the first thing to talk about, right, is is a bot. We're looking at bots. We're looking at them to do a single function very well. Now in that example, we can talk about a set of three bots. And and these three bots right here have very specific examples and functions they're gonna do. So one of them we're gonna call is a call wrap up bot. And what that purpose of that is is when you think about it as an agent does, a conversation with a citizen, they call in, about, hey. I need to turn my, my power on, and I provide all the information. At the very end of that call, right, they they tag all the information into their case. Right? So it has it all documented. Well, in that example, if we have a bot, right, that's actually listening to that call and can take that and then summarize what was said between the agent and the citizen and then actually document that for the agent. Right? You're saving time. That's important against it. You know, time is money, etcetera. But more importantly is that time you're saving for that agent now allows that agent freed up to get on another either phone call or an email or chat or whatever additional engagement, right, with another citizen, right, which is now a lot, allowing us to tackle more. Another example is a scoring bot, right, which gives a supervisor, right, the ability to actually evaluate, in an agent skill set or an appliance in a compliance point of view, right, if if, for example, they've answered the phone correctly or have given the right information out. Right? It allows us to score things faster, which, again, provides more time back from a supervisor point of view. Or another example is a specific knowledge bot, which might allow, right, when an agent's looking for a piece of information, why they're having a communication with a citizen, that could be either email, chat, or voice, etcetera, where it's able to provide that content to them in a faster pace. Right? We're able to prompt them, right, by listening to the phone call and providing some information back of what they've heard. Hey. Maybe this citizen's looking for this and and prompt the agent to have that information quicker, and you save time. In this example right here, right, if you're using these three, again, individually unique bots, right, to apply into, you know, an organization that has about two thousand agents, two hundred fifty supervisors, and about an average of an eight minute call duration, and you're able to save, for example, sixty seconds in your call wrap up bot, fifty fifty percent of your time less in scoring, and about thirty seconds less on a on a interaction. Right? From the knowledge bot, that can actually drive upward of fifteen million dollars in savings, you know, in less than six months. Now you start talking about how great that is. Right? And, again, this is not replacing physically, in this case, agents. Right? It's not replacing a human. Right? It's adding and supporting. It's augmenting. It's providing additional support for that agent in this case to support a citizen. And these are great examples. Right? Yeah. Speaking of great examples, Caleb, you have something? Yeah. Yeah. Well, I I just wanna add on to it because you're you're calling out something I think is really, sometimes it gets lost, I think, because these are great data points. And, you know, the the fifteen million in savings is is massive. I think, as you know, Scott, I mean, TTEC, we operate contact centers all over the world. We work with some of the largest agencies, you know, largest customers handling millions of interactions on a daily basis. We've been testing some of this technology in our own contact centers for the last, year. Since kind of this emergence of Gen AI, large language models, and especially on the summarization and on this, the scoring bot side, we what we've seen is, yes, there's a material impact of making things more efficient. But there's also just a a human element of this, which is you're just making this employee's life a lot better. So if you think about just this idea that I'm not having to sit there and physically write out three to four minutes of notes after every call, I'm able to let something do that for me. I can actually read what it put together. I can process it. I can edit it. It's just a more effective way of reducing stress from that job. And anyone that's worked in a contact center before, I think you know it's probably the most stressful job there might be out there. Right? Just handling those interactions. And then on the scoring side, one of the things we found is also from an employee retention standpoint. It actually is quite difficult to teach somebody how to go evaluate, coach, score somebody on how to deliver better citizen experience. And by leveraging some of this technology to actually proactively, you know, automate the the scoring of some of these things and evaluate and kinda give you a pull somebody else pull out kind of that human part of it, which is really, really important, as part of this evaluation of how you can leverage AI automation in this. Absolutely. Those are great examples, and and you're spot on. Right? That's the whole point. Right? We talk about it, right, and I emphasize it all the time. Right? It's it's about augmenting and supporting. Right? And and those are great examples of how that's supporting agents. Because at the end of the day, right, they are, the most valuable asset, right, that that you have right there that that's interacting. And as you've seen and we see all the time, if the if the agent's happy, right, by nature, right, their communication with the citizens or communication on the top Yeah. Is is is improved. It's happier. Right? Positive feedback and driving it. And, again, when you're scoring more and you're evaluating more, right, then, again, it also turns into things of, as you're just mentioning, it's not always negative feedback you're providing. You're providing positive encouragement feedback of things that they've done extremely well. You add in other bots that allow them to actually also improve and get on the floor quicker. Right? So when you start adding, you know, you know, other coaching or other real time, agent assist concepts in, right, and, again, this is using some AI and other technology, right, or other bots. But, again, now you're also enabling back to the point, enabling that agent to be more productive sooner, which also gives them more confidence, allows them to be, you know, in a better spot feeling like they can accomplish a lot. So there's lots of positive that get in when you start looking at how you're gonna apply your AI or or process automation into it. But it starts from the basic. Again, going back to you, you gotta understand where those datasets are. You gotta understand what some of those challenges are so you can apply it the most effective way. You know what? And one more example, what we'll do here is we're gonna I'm gonna pause, and we're gonna start this video. And and, again, when you look at this video, it's gonna give you some real time or some real live examples, right, of how AI can be interjected into, interactions with a citizen as well as with an agent and how it can support both. In this example we have here, this video was originally kinda more geared towards enterprise organization, but you you can apply the same concepts clearly in a public sector because there's a lot of similarities, as we know. Definitely differences, but a lot of similarities. So let's go watch this video, and then we'll come back. Hi there. I'm going to show you how a whole team of Verint bots augments the human workforce with automation and helps brands elevate CX. We're creating a workforce of humans and bots working together to increase CX automation. I'll show you several examples of humans and bots working together. Let's start with the Verint containment bot. Thank you for calling Acme Limited. I am Ali, your virtual assistant. I can help you with a variety of tasks. How may I help you today? I would like to confirm an order. I can help with that. Please tell me your password. Two three one one. Thank you. Miss Jones, I see a recent order that was placed last week. Is this the one you are inquiring about? Yes. I need to make a change on it. What type of change would you like to make? Change the quantity from ten to twelve. I can do that for you. You are all set. Thank you for calling. In this scenario, the customer's request was fully contained by the Verint bot and never has to be transferred to an agent. The Verint containment bot is part of the Verint open platform and can significantly improve your containment rates on voice and digital channels. Now let's take a look at another scenario where the call has to be transferred to an agent. Thank you for calling Acme Limited. I am Ali, your virtual assistant. How may I help you today? I'd like to change the payment due date. You need to modify your payment arrangements. Is this correct? Yes. Certainly. One moment while I get someone to help you to change your payment due date. Thanks for calling Acme Limited. This is Natalie speaking. Miss Jones, your identity has already been verified, and I see you need to change your payment due date. I'd be happy to help you with that. As miss Jones was being routed to the relevant agent to handle her request for payment change, two other Verint bots were automatically engaged. The Verint risk scoring bot generated a low risk score, which was automatically displayed to the agent as green, meaning low risk of fraud. And the Verint transfer bot uses generative AI to summarize the interaction so far and displays the summary to the agent for context. These two bots working together saved thirty seconds that would have been spent by the agent asking miss Jones authentication questions and asking for the purpose of her call. Miss Jones' experience is also elevated because she did not have to repeat herself. I can see we emailed you with some payment options. Did you want to go forward with the thirty day extension? Yes. I'd appreciate that. The agent was already aware of an email conversation from the prior week and therefore was able to confirm a thirty day extension. Now let me introduce the Verint knowledge suggestion bot. This bot has been listening to the call in the background and automatically suggests an article to the agent about how to change the payment dates, saving an additional thirty seconds of manual searching time and avoiding miss Jones waiting on hold for thirty seconds of silence. Miss Jones, we extended the payment date by thirty days, and you are all set. I see that you also recently increased the quantity of your order from ten to twelve. Many of our customers who purchased this product were interested in an accessory as well. Would you like to hear more details? What you saw here was the Verint coaching bot in action. Based on positive customer sentiment, the coaching bot guided the agent to upsell with an offer that matched miss Jones' needs. The Verint team of bots increases CX automation for Acme Limited. This results in increased agent capacity and more delighted customers and can turn a customer service call into a revenue generation opportunity. The call comes to an end and the customer hangs up, but the agent is not done. Now it's time for the agent to write a call summary and post it for the record, an activity which can take the agent sixty seconds. At this point, the call summary notes are automatically generated by the Verint wrap up bot, further increasing agent productivity by sixty seconds per call. We saw how AI powered bots augment your human workforce to seamlessly increase CX automation. The total call time was reduced by two minutes while also reducing agent effort and improving the customer experience. This scenario leveraged these Verint bots. With a large team of Verint bots supporting the agents, the opportunity to improve CX automation is significant. In our demo, one call about a quantity change was fully contained by the Verint bot, and another call about payment change was routed to a human agent. This agent was assisted by a team of bots to reduce the call length from six minutes to four minutes and create a more positive customer experience, which led to an opportunity to upsell. Given the large volume of customer interactions, brands can save millions each year with the Verint open platform, orchestrating a workforce of people and Verint bots working together to elevate customer experience. To learn more about how Verint Bots can help you drive CX automation, visit verint dot com today. Caleb, what'd you think about that video? I mean, it's a it's a perfect real world use case here to be able to actually apply, technology to make things more efficient, be able to actually, deploy this at scale, which I think is a really, really important piece of this as well. So, great example, Scott. Right. And and it's exciting. Right? I love where we get the bots, talking and and actually bringing it all together and actually walking it through. I think it's great. We have a lot more of these kind of videos if you ever wanna look at them, but but they're really great to show real world examples of how they can be applied. And what I bring this back to now is going back to the original piece. When you think about this, we talk about what that original challenge that we're really trying to resolve. And, again, it goes back to that engagement capacity gap where we know it's gonna continue to increase. It's not gonna get any better, meaning the expectations are gonna continue to move, higher and higher. We can't just throw additional resources out, as we said, to resolve the problem. And you can see, though, if you apply AI, bots, process automation intelligently, right, you're able to decrease that gap, get it back into a point where you can manage it, achieve, right, CX, the citizen engagement as well as agent, experience, you know, satisfaction, citizen satisfaction, pull it all together, and and really do some good across the board. Yep. I'm gonna pause there. And always, what we wanna do is think about takeaways when we wrap up these kind of webinars. So I'm gonna bring that up, and we'll talk that through. And, Caleb, we'll add a few pieces here. But, you know, one of the things we always talk about is is understand your data. Right? If this is something you're interested in, you know, when you're looking at AI, you know, understand where your data is, talk about that first, apply that. Also, understand some of your business challenges. Take a look at that and understand your gap of where you have that at. And, Caleb, is there anything else you want them to think about when they walk away? Yeah. I I think the big takeaway here is and and I can't emphasize this enough, which is the being super organized in the very beginning of this evaluation of where can I go apply AI, where can I go apply automation? And so one of the the ways that we found this to be the most effective is, just simply sitting down with, you know, people like Scott, with TTEC, with Verint for a few hours, and kind of understanding what are what's that technology stack that you're using today? Where are there opportunities to be able to go in and bring in automation and kinda map that out? And what we find a lot of times is there is some low hanging fruit places to go apply AI initially to be able to go get immediate benefit, but then also understanding that foundational thing that you can build on over time and be able to be able to continuously kind of in an agile model bring in new functionality, new areas of self-service, new places for these bots to be, to bring efficiencies. And so, you know, that's always my recommendation to customers is let's sit down for just a little bit of time. Let's do it in an organized way. Let's map it out, put a plan in place, and then go execute. It's a great way to do discovery with each other, kind of kind of get deeper into the tech and some of these use cases. So great. Thank you, Scott. Absolutely. You know what? Great examples. And, again, you start anywhere. That's the greatest part about this. Right? It's not a you have to start point a to get to point c. You can start at any one of these points, right, and and still get, value out of it and get benefits out of it. It's all about where you wanna start. I really appreciate you guys joining. Again, Scott, Caleb, from us to you guys, thanks so much. I will before we wrap up, two things. One, there is gonna be an ability for you to ask any questions, obviously, throughout this, and we will get back to you with any questions you've had. And then two, as I mentioned in the very beginning of this, there is, this is a three part series with TTEC and Verint. The third one's upcoming in a couple weeks. So make sure if you, if you've seen what we're talking about and you like it, register, join, participate in the third one. Lots of information we're trying to get out to you guys, and, we really appreciate you, attending today.