Name: Preparing for the future of customer experience in the age of AI
Uploaded: 2025-11-13T19:32:00.858Z
Duration: 59 min 48 s
Description: Preparing for the future of customer experience in the age of AI

Transcript for "Preparing for the future of customer experience in the age of AI": Hello and welcome everyone to today's session. My name is Nicole Duarte and I am one of the Community Managers over at Customer Success Collective. It is with great pleasure that I introduce to you today's panelists. Without further ado, I'm going to hand over to them for a quick round of introductions. Thank you for tuning in and enjoy today's session. Thank you, Nicole, and we're really excited to be here. My name is Dylan, and I'm the COO at Quack AI. I'm very, very excited to have this amazing panel with us today. So without further ado, let's start with a quick round of introductions. Carlos, let's go ahead and start with you. Yes. Hey, everyone. My name is Carlo Shearer. I'm the senior manager for the solutions and support engineering, team here at Hologram. We are a global cellular connectivity provider for IoT devices anywhere from e scooters to medical devices, industrial asset tracking, etcetera. And my teams, handle pretty much everything, post sales, customer onboarding, customer support, expansions, renewals, partnerships, etcetera. Perfect. Thank you, Carlos. Hey, Gal. Let's hear from you. Hey, everyone. I'm Gowling Geed. I'm head of customer experience at Faye. We're a digital travel insurance company that redefining what protection feel like. We're offering the best insurance, travel, and financial solution for travelers on their own, and we're blending smart technology with real human care. So we can make sure that they have they have their peace of mind during their traveler. My team is basically facing the entire interaction with the customers from the moment that they are making an interest in our product and until they're coming back home and need to file a claim. Amazing. Thank you, Rob. Hey, Jason, you're up. Alright. Hi, everybody. My name is Jason Barclay. I'm the director of servicing and collections at Notable Finance. We provide financing for people who are looking to sell their home and, looking to make improvements to their home before they sell it. So we partner with a lot of brokerages and agents and, contractors and, to be able to provide that service and financing options for for those who are looking to get the the biggest value out of the their home sale. Awesome. Thank you. And last but not least, Hailey. Hi, everybody. Happy to be here. I'm Haley Johnson, and I'm head of sales enablement at Artlist. Artlist is a creative technology company, so we're offering cutting edge AI tools, and access to all the latest, AI models for image, video, voice over creation alongside, the highest quality creative assets. So we're enabling creators to be able to stay on trend, achieve their creative goals, and very happy to be among such, amazing people Awesome. And as I. mentioned, I'm the one. I'm the CEO at Quack AI, and I'm also very fortunate to call these amazing thought leaders, customers. So at Quack AI, we provide, trainable proactive AI agents for customer support and customer experience. And the goal today is not to explicitly talk about Quack AI, but to get a lot of interesting insights from all the people in this amazing panel. As you heard, we have a very diverse set of individuals, but also very diverse companies and industries. So, hopefully, we'll have a very thoughtful discussion, and we'll get some insight into the future of customer experience and customer support in the age of AI. So let's jump into it. We'll start with some of the biggest challenges that you faced as customer demand, kind of grows rapidly. Obviously, depending on the type of business and and industry, but I think we're all seeing, demand increasing. How do you deal with those types of challenges specifically when it comes to, incorporating AI and different technologies? And maybe we'll start with, Carlos. So the village the biggest challenge we were handling was around, besides just volume, the technical variance between customer outreach. There were very repetitive questions that could be easily answered, and there were other more complex questions that required actual technical debugging. And we needed to invest more time on those more complex technical questions. And the amount the high volume of tickets that, came in that required simple answers, repetitive answers that could be retrieved from our knowledge base, limited our, our time that we have to address the other more technical complex questions. So we needed something that allowed us to quickly, go through these, more simple questions so that we can focus on, the more complex questions that we need. Great. Thank you. Jason? Yeah. I would I would, you know, piggyback off of what Carlos said. Right? Like, when we started to scale and grow and see volume increase, you know, the the first thing, that, you know, came to mind for me was, like, is my team gonna be able to maintain this amazing high level of, you know, customer satisfaction and experience that we've been, you know, we pride ourselves on for such a long time. And, you know, as as I think the pressure mounted on the the volume on the team, we wanted to to find a way to be able to allow our team to continue the that high standard, and, you know, make a resolution for the simple repetitive questions that could, you know, be answered quickly like Carlos mentioned. And, honestly, I think that's even a better experience for our team because even though our, you know, average response time was phenomenal, it it it still doesn't beat, you know, AI response, like, that gets sent immediately, to answer a simple question that somebody's asking for. So I think that was, you know, the biggest challenges for us. Awesome. Kelly, go. I'll take the next one for both of you, one at a time. So was there a specific defining moment when you understood that customer experience had to kind of, like, evolve or move from just, you know, ticket in, ticket out, just resolving to more of a holistic type of strategy? Sure. So happy to to start with that. I have to say that from my earliest days at Faye, it was clear to us that SIX couldn't be just about resolving tickets quickly and efficiently. We stable FAE with different approach. We know that to deliver six star experience, we have to have more holistic data driven and proactive approach. But I think that the moment it's really drive us to understand that we need to shift. Our approach was about two years ago during our insane, intense summer period, while we also experienced, a global outage of Microsoft that affected thousands of travelers. And in this breaking point, we understood that the old reactive approach was not gonna hold up anymore. And if we want to continue to maintain our six star experience, how we love to call that, we have to change our approach here. We need to make sure that we are able to provide our agents, the free to to really manage those complex cases and where our empathy and our reassurance are more, crucial and how we're able still to be proactive and maintain these high customer satisfaction and make sure we're taking care of them during their trip in real time. So that's where we start to shift it all and made to the added to the end to end experience at anticipate basically customer needs, blending the emotional excellence with emotional intelligence as well. I'll jump, jump in, there for us. So, I've always found it such a privilege to work with creators and in an industry where our strap line is, like, enable people to create without limits. And so when I first took the, kind of reins of, looking after our customers, we were able to do that. You know? We were able to service, their needs, and we could give them a real personalized experience with a human to really kind of focus on, how, we can get them, you know, focusing on the thing they love, which is creativity and not having to, be in kind of long support queues, etcetera. As we started to scale, it became more and more difficult to be able to maintain that. And, you know, you you start to realize that you've got to be able to adapt and move much more quickly. But we didn't want to lose the personalization, and we were tempted kind of by how how do we automate more, how do we try and, like, you'll hate me for saying this word, Duran, because it's your most hated word I know, but deflect Not not deflection. Not deflection. tickets. And it was you know, that was the place that we first started to look, and then, you know, we had to remind ourselves that that that's not what we're here for. It's not why we exist. You know, we don't wanna deflect. We wanna help. We wanna we wanna make sure that we help, in a way that, matters for our, customers. So and I would say we're say we're still on that journey. You know, there's, we're still kind of at that, point of being able to make sure that what, we implement and how we use AI within our service focus must not lose sight of, the experience that, the user has the minute that they are interacting with you. So, yeah, it for for us, it was how can we scale, deliver fast, efficient service, but not lose the personality and, the, warm feeling that they have of interacting with the business that genuinely cares about their creative needs. Yeah. Deflection is not one of my favorite words, but I I love the approach. By the way, Gala, I don't know if you saw, but Jason said he's stealing the six star experience, which I personally also love. So, so deflection, which, again, is a word I hate, is a kind of like a good segue because, arguably, CSAT is becoming maybe less and less relevant or just kind of like a worn out, KPI to track. So, Jason, this is for you. Like, beyond CSAT and maybe some other traditional KPIs, are there any other signals or behaviors that help you and your team kind of keep track of, whether customers still feel that, you know, genuine human connection when they interact with, you know, AI or the team or a combination of AI teammates and human teammates. Yeah. I think the big one for us because we were in the, you know, the financial industry is that, it's it's extremely important for us to track, like, any complaint. Right? You know, where our team our team is trained on it. It's required, you know, for different regulation purposes. So reviewing that, we look at it on a weekly basis, just, you know, leadership, all the way up to the CEO. Every week, we spend time looking at any complaints that would come in, to to see if there's room for growth. And we've had great ideas for improvements not just to our products, but, a good, you know, a good indicator if there's anything, that we're doing or maybe the pressure is mounting, that that may be a good indicator that we need to check-in and see if if we need to make some changes. So, that's that's one outside of the CSAT score. And then the other one is just really checking in with your people. To be honest, like, it's are are, you know, the people that are on the phones, the front lines, like, getting the vibe on, you know, how people are are are talking to us, not just solving the issues, but, you know, are they are they really feeling, you know, heard? And and are we still getting the good compliments that we've gotten about, you know, the speed, the answer, the empathy, you know, the emotional intelligence that our team is showing in when dealing with different circumstances. I think that's also a huge one, really, just to have a pulse of, are you listening to the people on your team that are that are dealing with your customers? Awesome. Thank you. So let's do a quick zoom in into Quack because, as I mentioned before, you're all Quack AI customers. And just as a reminder, as I mentioned, Quack AI, we provide, trainable, proactive AI agents. We like to call it an army of AI agents, so you can train a multitude and unlimited number of AI agents for different use cases and deploy them wherever you want to. So I would love to get everyone's take on this because I'm curious, and we haven't rehearsed. So I don't know what you're gonna answer. So I'm I'm genuinely interested. So when you made the decision to start deploying QUAC, and we know that there's a lot of options out there, a lot of, tech stack available, how did you balance how did you plan the balance between the automation or the Agenca AI experience, with continuing to maintain the service quality and standards that you wanted to upkeep for your, for your brand, the tone of voice, and everything else? And feel free to take it, freely. Like, whoever wants to take it first, go ahead. I can I can talk a little bit about that? So I did do a lot of research of different options and I tried other tools before, getting to implement QUAC. But, basically, the first thing I understood after my trials and investigation is that, you cannot expect just to plug in an AI model and expect it to have all the answers. Right? It's like you're onboarding a new agent, and you it needs to go through an onboarding and ramp up and QA and training and, you know, it's repetitive cycles of of continuous improvement for it to get to a place where you're comfortable on letting it running out of pilot, basically. Right? So, the way I did it was, first, instead of doing actual customer responses, it started replying on internal notes on the tickets. Right? And then us agents got to review, correct, pass feedback, train, or just, like, pass on the reply. Right? And from there, we once we launched it, like, actually productively, to reply to customers, we started noticing knowledge gaps. Right? Like, okay. Yeah. Maybe, where it's like, how to use certain parts of our product, it doesn't really know and it tends to escalate them more often than others. So we would focus on creating knowledge based articles and, going through training. cycles and corrections for those specific topics to improve those topics. So, it's it's always improvement. Right? And it's exponential. The more the more you work on the model, the more it will, be able to run on its own and the more you will be able to focus on more technical tasks or on, more, enablement both for self-service customers and the AI. Yeah. Thank you. I I Jason and I had a fireside chat, I think, about a month ago, and I love that he said it's, I think something along the lines of it's not plug and play, and that you basically I think it was like quack gives you the car and the keys, and then you need to decide how fast you wanna drive it or how you wanna drive it or where you wanna drive it. And I think it's something that there's some misconception out there that agentic AI or AI agents in general is just something that you deploy and then it's like magic, and and which is not the case. The the internal notes is super cool. I actually didn't know that this is something that you use because, we have a lot of support and CX folks, joining us, and will they will hear this recording as well. And this is what I used to have, and I'm sure a lot of other people also when we trained, human teammates that joined. Right? Initially, you give them a a training. They do some shadowing, and then they start answering only in internal notes. You check them, and you make sure that the quality is there. So it's it's cool to see that you used a similar method. Cool. Thank you. Whoever else wants to take the same question, go ahead. Want me to jump in? So firstly, Duran, I can't tell believe that you're, telling me that all those messages I get every day that tell me that AI is plug and play, and by tomorrow, I'll have an amazing AI agent. Like, you shattered my dreams. So, but I have to say it's not true. So for any of you listening, if you are getting bombarded as I am with, everybody and their brother and sister that's telling me that they've built the next most amazing AI agent, like, be cautious. Because one of the things that I would say, you know, that I'm so pleased that we did was that we took our time. And so, you know and and it's very, very difficult to take your time when you can see the service, levels that you've protected for such a long time are starting to creep, and you can see that there's this real pressure that you need to act and you need to do something. It's easy to make the wrong decision. And so we spend, a lot of time being able to really test different models and compare accuracy, before we, decided on, Quack. And once we decided on Quack, it was such a, phased approach in terms of implementation. So we took our lowest volume of tickets, for a product group so that we could do a really controlled test. So we put it on one, brand. We didn't put it on the leading Artlist brand, so that we could test it, and we could really kind of put it through its paces, in that sense. And even though that was a low ticket of volumes, we then only put it on two or three contact types and and really kind of perfected that. So I think the fact that, you know, we were able to, have confidence when we were putting it in the hands of our customers that it had been through thorough testing, and we could really iterate and learn. And the thing that we really started to fall in love with with Quack was the fact that we have this ability to control, some of the, kind of rules and responses and policies that sat behind it. And you started to realize how important it was that, you know, customers have that level of, of protection. And similar to Carlos, we have the human in the loop for as long as we needed to until we were confident, that we could, hand that over. And, yeah, we're now at a really exciting place where we've seen the value of it in a controlled environment with our motion array product and looking forward to being able to see what we can do. Now we can put it on the Atlas product. Amazing. Thank you. Well, I jump in now. So as I run, I think before we started, with Squawk, we start really smaller, by trying and kind of developing an AI tools in house. It will help us with some repetitive tasks such as, our documentation or other manual task. But when we understood it, we want, to expand the Gen AI and include AI as well in our front line. At the beginning, we test a few different, companies. And what we, what we focused on in our POC is evaluating the response that we're getting, how accurate it is, but more importantly, the empathy and, how personalized in humans can be. That's how we end up, with Squak. Not only did they have the best performance, but also we had, one sub show where we can have everything in one place, including the Geni, including the quality assurance. And as Kayla said, to have everything, visible to us in self-service, platform, which make it much more, easy. So at the beginning, we were still very passionate about releasing our Gen AI, to customers. So we did start small second there. We took our one of our lowest channel and actually potential customer where we have more general inquiries, and that's really led us to the ability to test the tone without compromising our experience and really does it slowly slowly until we were able to expand it. Also, in the early days when we did it, we start with, first of all, mapping our knowledge base gap and making sure that Quack is up to date and you have the best information and knowledges that every and any experienced agent have. And when we did launch it, we did see the human, human being reviewing every interaction at the beginning to ensure that the empathy and the accuracy, is as we expected. So when we saw that this even when it's not treated in the way that we want to, we're able quickly to train him to do differently, to adjust, the training that we gave him and to be able to increase not only the deflection that around us don't like that for us to use, but also, the customer satisfaction and really, make it visible like an fake agent talking with them. Awesome. Thank you. I think we're gonna go into a slightly sensitive topic, and we'll go into addressing concerns, specifically internal concerns. So it could be your team or it could be, other teams potentially about AI replacing human agents or replacing humans in general. And this is I know it's a touchy subject still. We, at Quack, really believe that AI are teammates. This is what we call AI teammates versus human teammates, and we genuinely believe that AI is not there to replace humans. You know that this is also the way that, we deploy the technology and also the way that we, recommend your teams to do it. And our CX 2028 initiative also talks about it, and it was really cool, actually. We just had a master class in in both London and Boston, and we're doing another one in London next month. And one of the trends that we, forecasted for mid twenty twenty six was new roles emerging. And then I'm not gonna mention which of the the companies of of the attendees here, actually posted, that they started as one of the positions that we envisioned for mid twenty twenty six, and also a different company this present here. So, number one, it's really cool to see kind of, like, old school positions within customer experience and support and customer success, transitioning and not being replaced. Just new jobs are being created, new positions are being created, and we kinda talked about it in some of the quick internal webinars that we've done. But my question is, how do you address those concerns, like, stemming from your team or from other, stakeholders, and kinda shift the narrative towards, like I said, empowerment and showing them that they're gonna have other growth opportunities as AI becomes more prominent into specifically CX and support. And, again, feel free to jump in and and take it as you please. Well, No. Go ahead, Scott. yeah. Okay. So I think first of all, communication is the key here. From the first day, I always told my team, AI is not here to replace you. It's here to interest skill, to make you, again, your work more efficient, and free you to work on more complex and interesting interaction. More than that, it's also gonna open a professional path. So but the one thing that we did, we we didn't start by introducing Gen AI directly to the agent and and that. What we did was, first of all, to use an AI tools. I talked about it earlier. We tried to remove, first of all, the repetitive manual work such as documentation or how to get to some of the knowledge easier. And when they saw how actual, their job make how to make it their job easier, then the mindset was naturally shifted from fear to empowerment. So when we did launch our Gen AI, our team was very excited about it. A lot of them actually wanna take part of that, and they are continue providing feedback on that and see how they can today also use AI to be more empowered in their day to day work and overall to provide our better customer experience to our customer. So I think it's really how we're communicating them to them. And once they get it, it's making it much more easier. I agree that communication is key. Always making clear that AI replaces tasks and not people or not roles. In our team, instead of having people copy pasting knowledge base links or API instructions or SIM activation instructions, we let that for the AI automation, and then our team can focus on, not only more technical troubleshooting, but also other, strategic initiatives like where we just revamped and automated our customer onboarding process. We are taking a a lead on the gross margin initiatives. We are creating new automations, scripts that will help us internally and our customers. So our team gets to grow, you know, technically, develop their technical skills and focus on more technically complex roles. And we, we allow the AI to, you know, handle, the more simple tasks. So, again, it's not replacing roles. It's enabling the team to grow and do more complex tasks. I love that. I love it. I can I can jump in here too as well? Like, one one big piece for me that's always been, like, a main focus, something that has been really important for me is, you know, career development, when it comes to, our people. You know, especially, if you've ever had a time where you've been like, man, I wish I had more time to give you that experience or if you had people on your team say that to you. You know, this is one of those things that, is going to be able to not only help you scale and grow as the volume increases, but, you know, also it can come in and help you free up that time so that you can invest in the the things that you find value in and invest in your people and help help your team have more time to do that career development and growth. So, yeah, like like Carlos said, I think, you know, free freeing up your your time for your team to be able to focus on more technical things and and the things that they actually can put their talents towards, that are important is is a huge is a huge benefit. I can probably add just another thing on there, Duran, in terms of some of the other departments kind of that get excited, about it. So, before having, a tool like Quack, we had the traditional kind of, help center, and, chatbot, like, rules flow, etcetera, that was in there. And it always felt like a bit of a battle in terms of, being able to get investment from, marketing teams, design teams, etcetera, in supporting, like, your help center content. You're, like, being able to kind of have all of the right materials that are, are needed in, like, the support workflow. But I think being able to use, Quack and the implementation of it, you know, a much more agentic AI experience at a time where I think the whole world is watching as to, how AI will reshape the way that we, live our lives, let alone kind of handle our our customers, was that we could actually get some of the departments that perhaps hadn't been so involved previously, involved from the start. And so, you know, you'll remember when we did our first test day, you know, we did it live in the office. We had people from product, people from marketing, our CEOs, like, that you know, we had people there, as part of that kind of initial, test to be able to see it from the start. And so, I think that's been the thing that's been great for us is, you you know, getting more and more people involved to be part of, the journey, the learning, etcetera, and then being able to see the benefit that they actually get from the reduced released capacity. So for us, it was never about reducing headcount or saving roles. It was about saying, look. Let's free up these, people that are the closest to our customers that, you know, understand more than anybody in terms of what our customers, need. And rather than having them answer and repeat simple questions, let's take that amazing talent we're lucky enough to have and put them in marketing, put them into, our data teams, put them into the product teams, and, you know, being able to have a tool that is actually releasing capacity made that possible. And so it's been amazing to be able to kind of see feel the benefit that we've created roles that I don't think we would have had the opportunity to do, had we not, you know, reaped the benefits of, some of the efficiencies we get with Quack. Perfect. So this is a great segue. So I'll direct this to Hailey. You can continue. And then Carlos, I would also love to get your take on it. So looking at what we just talked about in terms of, you know, the concerns about AI replacing human agents and the shift and how do you, how do you shift the narrative towards empowerment? How did deploying quack or evolve evolving or emerging technologies kind of shifted what it means to be a customer experience or support agent in terms of, like, the r and r's, then the role, and the definition, the day to day, maybe the purpose of how, you know, those team members view themselves. I think it's a great question. And I think one of our, biggest reflections when we first started to, see the benefit of QUAC kind of picking up, like, let's say, the first response, and, you know, doing some of the initial troubleshooting steps, with our, customers is that we realized that we had created a team of human chatbots. So in reality, when you started to take away a a simple macro that, you know, would would provide the agent with the answer, because Quack was handling that, you you realized actually there's a real gap here in some of our, team's capacity, and the tools that they have to be able to actually think about a really unique and and personable answer. So, it's you know, the first part was reshaping, the role that they play with each interaction and and knowing that, actually, it's not about just closing a ticket. It's about stopping and thinking and understanding actually. You know, there isn't necessarily a simple copy and paste answer to this nor do we want there to be. So I feel like the first thing was that it felt quite, exciting and empowering for the team to to be able to say, oh my god. I've not got macro anymore. I've just gotta think about the right answer and actually have a, an enjoyable conversation with somebody. So I think, like, the core role changed and became more enjoyable. And then we started to see new roles being created. So we hired, in, experienced resource, to manage the, implementation and somebody that had already had, you know, a a year or so of implementing these tools. So we had people coming to show us, like, how to do it and teach us new things, which was brilliant. And then the team started to wanna get involved in those things too. And we were able to start to create dedicated, roles, which then starts to give you career paths for people to see, that there's more of a career in support as opposed to I think, traditionally, perhaps support roles are a bit of a gap role, a filler role that, you know, you'll do for a few years, and then you'll you accept that you're gonna leave and go and do something different. Whereas I think, you know, it's really kind of changed, the definition of what it means to be in support and, you know, that people can actually see a a career, as part of that. And I think the fact that we have technology overseeing every single conversation that is happening and using AI to be able to analyze and understand and group that together and start to track trends, the team is so excited about what they can now do with that information. So, you know, they they can truly represent the voice of the customer now because they've got the tools to be able to actually rather than just hearsay or judging what we need to do in our product road map based on the last conversation you had, They're the ones that are owning the data that say, actually, no. I can show you here. We've had 50 conversations that are based on this, and, hey, let's delve in and see see what they are. And, you know, the fact that that crack is over the top of every conversation is a real, win for us in terms of the the voice of the customer, but it's the agents that are talking to the customers that own that data and and represent that with our teams, which I think is great. So Absolutely. Carlos? Before I get into how it's reshaping the roles, that piece you said about QUAC being on top of a conversation, it's something like I mean, QUAC does very well, but I think in general, by having this, agentic model is that, you know, customers reach out. If it's a simple question or even not simple, but something that the AI can automatically respond, it does. If it doesn't, then it escalates. Right? So what we're doing is for those that are that need escalation to an agent, our AI model is helping us, collect all the information that we need to start actual troubleshooting. Right? So it's okay. You need to run these scripts and share these modern diagnostics, these outputs, etcetera. So but the moment that one of the agents gets to ticket, it already has all of or most of the information that we need to actually start troubleshooting the issue instead of having to, you know, spend more time back and forth trying to get that information from the customer. So that's a a great point, Hailey, that I that I really like about this approach. In terms of how we're shaping the role, other than at least, you know, within, Holoram, we had, what we call the past sales customer customer experience, which was solution engineering and support engineering. Support was support agents, ticketing queues, etcetera, and solution engineering was more focused on strategic accounts, partnerships, growth, and more specialized projects. Right? About a year and two months ago, I took over both teams and merged them, and, I started reviewing how I could make support more efficient, because I noticed that our support agents were on that never ending battle, against, ticket volume. So, the, like, the the amount of, open tickets that we had at one time, like, it was really large and the agents didn't have enough time to go through all of these tickets and then you know, while more and more were being opened. And, that's when I realized that we need to start implementing some sort of automation because some of these questions, even though they were simple, they required agents to go to get the appropriate knowledge base link, macro, like you mentioned, and it was just more like a information gathering role rather than actual technical support. And, I didn't like that for them. They didn't like that from the it was not efficient and efficient than a customer experience. So that's kind of like where where where I started adapting. Right? We implemented, our first attempts of, like, automation AI, and we quickly reduced the amount of, open tickets at a at a given time for about, like, 70%. It was absurd. Like, 70% of our open tickets were closed within three weeks of implementing this automation model. And, then, right now, we are or I already deprecated the traditional support agent model, on my team. We are all solution engineerings now, and I'm doing technical enablement and workshops and growth and, you know, database skills, coding, scripting. I'm making our, what were our support agents actual solution engineering engineers that can be involved in those growth and partnerships and having those more meaningful, relationships with customers, helping them succeed and grow and expand. So their their role is, more rewarding. They feel it's more rewarding, and they get to face more technically challenging, more technical challenges and, you know, just, like, growth with the customers. The customer experience is better as well. We have more resources to spend on those strategic partnerships with customers. So that's kind of, like, how the role has evolved from a supporting agent to a solution engineer. Amazing. As I said, all of you are kind of fitting perfectly into our vision of the CX 2028, and it's amazing to see how it it is a futuristic type of road map, and it's crazy to see that everything is, like, you know, just coming way ahead of how we planned it. Like, things that we planned to mid or late twenty twenty six, we're already seeing now. I wanna throw a curveball, and I wanna open it to everyone and whoever feels comfortable, jump in. If not, that's totally fine. And I just want you to use one word. Okay? So, I think it's evident that AI is becoming a core to daily operations. Right? When you are hiring now new agents or team leads or whatever the position may be, what is the number one competency or trait or skill that you're looking for? Just one word or one skill or one, competency that you're looking for. Just one word. Curiosity. Love it. Adaptability. Man, that's a hard one. I think for me, it's the type of person I am. I think positivity. That's a big one for me. Yep. I will actually jump on, Helen, what you said with curiosity. I think that's curiosity Mhmm. with with self empowerment, I will say even. Those are the things that truly matter today in order to be able to do more. Yeah. What what I love about the common denominator is all of them are human traits that sorry. But AI is still not able to replicate, at least not one to one. And, again, this is something that I I just it's it's a question that I love to ask in every single forum, and I just, you know, I just wanted to quickly, throw this curve off. So thank you, and the answers are spot on. So you passed the test. So being hired. you're hired. Okay. I wanna leave time for q and a, and I see that the q and a bucket is really filling up. So I think let me take one last question that I want to cover, and then we will open the floor because, otherwise, we're not gonna have enough time to give attendees the time to to, ask you questions. Let's take This one is so many good questions that I wanted to cover. Okay. I got it. You could give one piece of advice to a c x support customer success leader just beginning their scaling journey. What would it be? Go for it. So my advice will be, first of all, start plan it in early stage. Don't wait for the scales and the volume to come. Second will be scale this intention, not impulsive or maybe the same thing even in in a different way. But I think that the one big, lesson that I learned is not to rush to automate or to or to go around the AI buzz word that we have now. And really, first of all, to understand in your customer experience support, what is, what is truly, making your agent making their adding their, their value? What kind of things, the AI won't be able to replace? And what are the kind of things that we would never want also to to give customer to speak with an AI. And by a good planning, understanding where are the frictions, where are the places that, again, we want or don't want to, let AI to be involved, I think that, we can plan it much better. And with that, you can build it smarter, more human experience as well at the same time, and scale faster. So plan it. Don't be impulsive. I I was gonna say patience, which I feel like you've you already kinda touched on, Gail. So, it's instead of patience, I would say, collaboration, with your team. Right? Getting everybody, as many as many eyes, you know, on on it as possible because, you know, you you have your best people, but, you know, everybody's good at different things. And if you can take the strengths from, you know, the the great people that you work with and and use that, and combine it, it's just gonna make, you know, make the process and and and make your overall success, you know, I think better. So I think for me, it's like start with your why. Like, you know, it's before you, jump into solution mode and implementation mode, why do you exist in the eyes of your customers, and what is it that you want them to, experience? What do they need to experience, so that you kind of have definitely got alignment in terms of, the kind of overarching purpose of the the service organization in your companies, like, why, in that respect, and then being able to have a reality check. You know, you've got to face the brutal facts of what is actually happening today. Get your head under. Lift the bonnet. Like, understand exactly what customers are experiencing today so that when you start to design the workflow, you're designing it on the the basis of fact, and how your customers actually feel and experience in things, not what you think is happening. And I think it's very easy as leaders to, you know, that you designed something a year ago or two years ago, and and therefore, you start from there. And you think, right, this is what I need to improve because this is what happens. But in reality, there's so many other things that are happening in practice. So, you know, same as Gail said, like, and Jason said that patience part of it is so important, but you've gotta be brutally honest with yourself about what's actually happening, to know that, you know, you're you're definitely fix it fixing the right things. You're not just, kind of tinkering with something that, you you perhaps haven't fully understood, the the impact that you could be having if you designed it, with outside in? You know, what what do your customers need? What are they experiencing now and come back from there? For me, it resonates a little bit with everything that's been said, but it's, you know, just like set your expectations. Like, again, don't expect it's a plug and play. And You know, you need to be patient. You need to make sure that, you you spend the time, you know, doing the work, improving it, training it, noticing gaps on your knowledge base, on your customer interactions, fixing those passing, you know, training cycles. It will make mistakes. You know, we've seen it give incorrect or inaccurate process. We need to go and clarify to the customers once an agent spots those responses that were not correct. I mean, it's not gonna be easy, but it will get exponentially better. And, the rewards for the time you will spend implementing and training and making sure it's accurate, they pays off, you know, down the road. So, you know, set your expectations and spend the time working on it, and it will reward you. Awesome. Okay. First of all, thank you. And we're very fortunate because, number one, we had a lot of participation in the live chat during the conversation, which was really fun. And we also have a lot of questions. So what I'll do, with respect to time also because we have a countdown of thirteen minutes left, so we'll try to cover as as much, questions and answers as possible. I will ask a a one question at a time, and then we'll just do first come, first serve. So if you feel comfortable answering, just, you know, jump ahead and answer and then the next person, and then I will switch to the next question. So, Alexis is asking, I'm interested in hearing more details about how, Quack helps analyze and track on non deflected tickets. So I obviously can answer that, but I would love to hear it from our QUAC customers. So whoever wants to take it, please go for it. Should I jump in? Please. So, this was one of the first categories that we, wanted to, Mhmm. group together and see. So, that's the first thing is that it's grouping the outcomes, I guess you would say, in terms of the what gets handed to an agent. So being able to look at that and say because, actually, you will realize that there's quite a few conversations that you want to hand to an agent by design. So the first thing that we wanted to understand is you've got, true self-service in terms of, the the customer could get everything that they required, and, their ticket was not deflected. It was handled. I wouldn't say deflected. The the quack, agent handled it peacefully, versus the quack agent couldn't handle it, and it then comes over to a human. And we categorize that to say it was handed over by by design. So, we set a policy in place that says if there's any of these scenarios or characteristics of the conversation that happens, we want it to get to an agent. And I think there was another question I saw, like, how do you make sure that you don't lose kind of that human touch and empathy? Mhmm. A lot of policies that we build make sure that we recognize a human's needed here. Let's get it to a human. So we understand, it's gone to a human because we wanted that to happen and that was by design versus it's gone to a human because it couldn't answer. And then those are the ones that we then analyze. So we then look at being able to analyze those. We have dedicated resource, that's looking at that every day to be able to see what why did it not answer. Was it because we haven't perhaps written the policy in the best way and there's something we could improve in that? Is there a gap in the knowledge base and therefore we can improve that? Or is this just a obscure, you know, scenario that we accept, you know, that it it's not gonna be able to answer everything? So that would I say it would be the best, like, jump that we had was where we could look at separating, that is escalate to an agent because we wanted it to versus it just couldn't answer. Yep. Anyone else that wants to take this? Sorry. I was muted, apparently. I'll jump to that and say we did the same thing, but I think the one, additional value that, we got from Quap was the ability to get, more deeper insights and the sentiment of our customers. So in our regular communication system, we worked with ticket, sorry, with Tega system, and that's really not reflecting the reality. It's depend on the agent as well, and it's not really giving you the core topic. Now with Quaca Analytics, we were able to dive into your understanding really what are the core type of questions that we're getting from them. And with that, to to able to bring more, interaction that the Gen AI can, the Quacka agent can handle, and especially the sentiment. It's helping us deduct exactly again into the largest when customers already upset or getting, you know, in front of the chat, and we are able also to, better, ensure our empathy and our tone and our voice are delivered to our customer in that way. Amazing. So my the next question we're gonna take from Fringes. I hope I'm pronouncing the name correctly. They're asking, do AI trainer roles require coding knowledge slash experience? Not necessarily. It's almost like training, human agents, just the process is different. You just need to be able to identify knowledge gaps, be good at documentation. So if you if you develop on your knowledge base, which fits AI in model, you know, you need to write proper guides, proper format so that a AI can understand it. Like, it's mostly around identifying gaps, understanding format, and how to fit it to the AI. The most efficient way of running training cycles, you don't I mean, at least on my CX team, we don't really need to have any, coding knowledge to train them. Yeah. I would say, Notable, it's just the same with us. Like, we don't have any we don't have dedicated AI trainers. I think it's it's been a group effort. We just the. the the people at your company that have the knowledge, they're the ones that, can can handle the training for sure. Anyone feel? like sorry. It, wasn't. I I feel it I mean, it kind of depends what what what, kind of questions you're receiving and what your product and your company does. Right? So in my case that we do IoT cellular connectivity, even though they work the training part, you don't you don't require the technical knowledge. In order to train the AI to reply to our customers, you do need cellular modem and cellular networks in regards you know, that experience in order to give you that correct information. But that's because of our product and not because of. the skills. needed to do the training. Right. So I will say from a maybe subjective, but, you know, an honest product perspective of Quack. And, hopefully, you all can agree with me as customers. The way that our, unique training panel is designed is to really treat training your AI agents the same way that you treat training your human teammates. Okay? This is a really key philosophy of of how we built our product, and the way you do it is literally just use natural language. We're actually working on also just being able to record voice, and then it will be translated into prompts. But even now, if you would like to provide training and briefs and instructions and prerequisites, and for lack of a of a better comparison, just like you build a custom GPT, you are able to use completely natural language, and then Quack will actually enhance what you, ask the AI agent to learn and do, into the most optimized prompt that the LLM or the model can then turn into something that will be executed and optimized in the best way possible. So in short, the answer is absolutely not. However, as Carlos mentioned, you do need you do need to choose the right SMEs or domain experts, within your team and department and CX organization in order to train specific topics within, Quack's training panel. Because at the end of the day, as I said, just like you would have a new, human agent, you would want the specific, you know, most knowledgeable other, more veteran, team members training them. So it's the exact same approach. Only you're training an AI agent. And, usually, AI agents, once you train them, it's basically one and done. You don't need to do repetitive sessions with them, on the same topic. You may need to do some iterations and give them some more information, and do some, some repetitive regression testing, but you definitely don't need any specific coding or, like, technical expertise. So we're running, out of time. We have five minutes left. Let me see if I find a specific question that I want us to give to you. Ron, whilst you're looking, I'll just jump in and answer the other. question. Please. that was, Please. a minute. ago. I. think it's. a really important one. Yeah. Do. you need to have, like, dedicated, roles for it? In that sense, like, I don't think this is something that people can do on the side of their desk. And we tried that to start off with. We tried small and. a few of us just had extra responsibilities to try and make it work. And it just doesn't. Like, I think there's so much going on in the world anyway. Like, if you're gonna execute something like this, you have to have the ability to narrow your focus, and you can't do that if you're responsible for so many other things. So one of the best things that we did was being able to create dedicated roles with dedicated measures of success. So those people, you know, had a real clear, understanding of what they were responsible for, and they were given the freedom and the headspace to be able to just completely own that. Mhmm. And, it was you know, it helped us to move so much more quickly, having that dedicated resource. So I would definitely. recommend start start before you really need it, to the point that was raised earlier. Like, say start start early and get dedicated results so that that is all they focus on and you've given them success measures so that you know, you know, that it's delivering what you're expecting. Absolutely. Okay. We have time for one more question, and then we'll wrap it up. So Chris is asking, is companies adopt AI to scale faster and serve more customers? How do you ensure the human touch, empathy, authenticity, and brand personality, doesn't get lost along the way? I think it's all started by, as we said, the planning. First of all, is to understand which cases or which interruption the AI should handle, and if you can feel, first of all, safe with that or or satisfied enough for that. I can tell you what is what we did. In early stages, we had, very a clear principle for our tone, empathy, and how we're communicating with our customers. And we even gave it, that we gave, even for the small details, attention such as our images or phrases that we as Faye using it. And as we said at the beginning in early days, we made sure that we have human monitoring every interaction even before we launch it to production, especially after we launch it to production. So while we are able to track sentiment and also evaluate the early every interaction in the early days, we were able to adjust the tone very quickly to ensure that it's it's, fit to our original, tone and, the culture that we are having here. We have time for one more panelist to answer that question if you want. Yeah. I was just typing in the in the chat, if empathy and authenticity and brand personality is already the standard for your team, Mhmm. and the agents you have, then I don't think you'll have a problem because, you know, you're going to be training your AI to to mimic and and and follow the example of your best people. And if that's already a priority for your team, then you're gonna train the the new member of your team the same way. Amazing. Well, first of all, I wanna thank all of you, Carlos and Gal and Jason and Haley. It's been an honor and and a privilege. I think some of us have been working together for well before quack, by the way, for probably, three years or more. Jason and Carlos, I got to know you more recently, but it's been a genuine pleasure to work with you and know your companies and know your teams. And for me, again, we haven't rehearsed any of the questions. So it was really, eye opening and really fun to hear, like, different opinions from, you know, various, leaders from different parts of the world and very, diverse and different companies serving very different customer and user base. So thank you very much for your time today. For our listeners, those are, doors that are live and those that would listen to the recording. We obviously invite you to go to quack.ai. If you wanna see us in action, we definitely encourage you to do that. And definitely go and visit cx2028.com, to kinda see our white label and how we envision customer experience evolving, until 2028 from back to, mid twenty twenty four. And, otherwise, thank you so much for attending today, and looking forward to the next webinar. Thanks, Thank. you, everyone. bye. Thank you. Bye bye.