Buong transkripsyon ng Operationalizing AI Inside the Enterprise with Nate Christiansen

accustomed to calling for any need that they have, you know, set back since the nineties. And so to go and take that experience and just yank the cord on the dentist and say, sorry, you can only chat with us. Yoni Leitersdorf: Mm-hmm. Nate Christiansen: That's not the service nor the experience that we wanted to provide to our customers. And so this was a change health care outage happened. I want to say towards the end of 2024. And so, you know, the beginning of 2025 is when we really began looking for what is a voice AI solution that could meet the customer where they're at, answer their issues in a comfortable, good way. And be available to them in multiple languages, any time of the day, and also make it easier operationally for us as a business so that we're not having to train our people on thousands of different skills and products. Yoni Leitersdorf: Okay. So and and I I actually like that sort of thinking around avoiding the just sending them over to chat. A lot of vendors do that. I find that sometimes frustrating that you're you know, you're used to talking to someone or used to some kind of service and then all of a sudden you get this this smiling chat that has this nice name. It's called, I don't know, Zesty or or Neo or whatever. Yeah, and and you're like Nate Christiansen: Yeah. There you go. Yoni Leitersdorf: I didn't want this guy. I didn't want this chat. I wanted somebody to answer my questions and I didn't want to try to, you know, figure my way through this. So first of all, kudos for you for realizing that early. So okay, so voice based solution. what did you do? Did you build something, try something, build an RFP, meet vendors, you know, what what did you do at that point? Nate Christiansen: No. Yeah, so I think w what's neat with AI technology is things are rapidly evolving and it's like every week there's a new competitor, a new player in the space. but back it seems like back in the day, but literally a year ago, there wasn't That robust of options. So we did actually implement a chat. We had Agent Force through Salesforce for their chat for AI. And that was helpful. It was just a supplement, though. It wasn't there to replace the voice. But then, you know, all these different companies, Salesforce, Intercom, about every company started saying, Hey, we've got our voice AI solution. And at that time, you know, as large as some of these companies were. Or are it was honestly slightly embarrassing how rudimentary their products were. And so we had to go vet all the vendors and not just take them by like face value as hey, your product is great. We had to really go and test it and test it with our support representatives, test it with some customers until we landed on one that we felt very comfortable with and that wasn't giving that Siri robotic. type of experience with a customer where you know you yell at it five times to play the playlist that you want it to play on Spotify. So Yoni Leitersdorf: Yeah. Or you yell human

at it ten times like human representative. Nate Christiansen: Yeah, or you yell human. Exactly. Exactly. Yep. So I'd say, you know, the the biggest thing was Determining who's a faker and who who's real and who's actually has the technology that they're claiming to have. Now, today it's I'd say easier. There's a a new one that's coming up every day and with the models getting better. I know Eleven Labs has a lot of really neat voices and things that you can use and play with. but I'd say the biggest thing was first we had to select the vendor that we were gonna work with here. And then second, we're fortunate enough that we have over 50. knowledge base articles for all of our products and how to train it, how to use it. So we tied it into our knowledge base and we were able to give it all the information that it needed to properly support our customers. Yoni Leitersdorf: That's amazing. Yeah. As we're talking, I'm looking through your guys' website. So, first of all, I love the fact that when you click on support, it shows you resources, but immediately below it, it just gives you a phone number. Which, you know, surprising, but not that common. You know, a lot of times you look around, you fumble around, you end up in a chat, and then you convince the chat to let you talk to a human. Nate Christiansen: Yeah. Yoni Leitersdorf: or to anyone actually. And then here it's you know, I can dial and of course the AI might be the first one to answer, but I'm also seeing all this content. So you so you basically said, Okay, we have all this content, we can train AI on it, and it can have the conversation that, you know, these reps would have had anyway. what was the rubric like? What you know, what were sort of the main criteria other than, you know, it's actually a real solution and not sort of a fake it situation? but what was Nate Christiansen: Ha ha ha. Yoni Leitersdorf: What was your rubric evaluating? Nate Christiansen: Yeah, I'd say, you know, a c a couple of the big things is we were really looking at we believe in that human touch and having a a good experience that feels human and relatable. And so the last thing we wanted to do was give the customers a robot that is just frustrating that they want to push through. And so we actually took some of these different voice agents, we yelled at it, we swore at it, we we ran into the ringer and we Yoni Leitersdorf: Mm-hmm. Really? Nate Christiansen: But we tested and even trained it on some like emotional intelligence. And what's actually really neat is like you could yell at me and scream at me the way that we yelled at it and it would probably damage or hurt me a little bit, right? Yoni Leitersdorf: Amazing. Yeah. Nate Christiansen: But it handled it so gracefully. And so that's when we knew, hey, this is the right tool, because it's it's recognizing the customer's emotion and it's responding in a way that is gonna de-escalate the scenario and help drive the customer to the outcome that we needed to

get them to. And so that was actually our number one criteria was that customer experience. The second thing was, you know, I would say Most AI models are a reflection of actually what's it's trained on and what's in the knowledge base. And so how easy was it to train and to make adjustments so that we could quickly pivot and adjust, you know, if we have a bug or another error, let's test just making that change and making sure that it's not hallucinating and it's quickly grabbing the newest and best information, which is really important to us because we're releasing new features on a daily basis. Yoni Leitersdorf: Mm-hmm. Mm-hmm. Nate Christiansen: and we need it to stay up to speed with with the most current things that are happening. Yoni Leitersdorf: Amazing. I it was probably fun by the way, trying to trip the AI, right? Yeah. Yeah. They're just like, I'm gonna go at it and see if I can get this thing to to curse me, to yell at me, to hang in my to to hang up the phone. Nate Christiansen: it was a blast. Yeah. It was a blast. Yeah. Yeah, and then I would I mean th there's a whole and I'd say the hardest part of this actually wasn't even training the AI. And we'll be talking about that in a bit, but it's the financial modeling, the building a true ROI behind this, presenting and getting executive alignment, and getting the whole business to really sign off and align around a tool like this. Yoni Leitersdorf: Mm-hmm. Nate Christiansen: But I I will say, like, we we did a lot of A-B testing. I mean, in what world could you take which we still have incredible agents and call center agents at our company that handle more technical s scenarios, but there's really not a world that I could say, Hey, I need you guys to learn all of this instantly, and we're gonna change our entire approach to how we support overnight, that they would be able to do that. But with this, like you Literally change it, see the results using like a VI type of platform, evaluate the the results. Sometimes it was worse. You know, we'd change the intro message and make it say, Hey, I have everything about the Ascend product and know it all on my fingertips, and that didn't respond very well. So then the next day we would change it a little bit, but like we could get feedback. on you know hundreds of calls in a matter of days to help make very quick decisions to drive improvement. Yoni Leitersdorf: Amazing. Okay. So first of all, you had the rubric, you put it through the ringer. You said, okay, you know, here are the solutions that make sense and they they they they will deliver the service that we want to and they we can manage them, we can feed them, we can do what what is needed. Let's talk about the finances. So how did you think about the financial model of this? Nate Christiansen: Yeah. Yeah, so it's really hard. and I would say even a year ago, it was much harder to really figure out how to tie the number of tokens or you know, I like to call magic beans back in the back a year ago because people weren't even familiar with tokens. So I'd go to the CFO and be like, Hey, this is gonna consume this many tokens, and he meant like, Okay, well, if I plant this many magic beans, what am I gonna get out of this? Like and so what we did is we've actually Yoni Leitersdorf: Mm, yeah, yeah, yeah. Yes. Nate Christiansen: And this is kind of the model and the approach I like to take now with most AI initiatives. We ran a pilot. So we said, okay, all I need CFO is 40k investment to give this a shot. Let's let's get a test scenario up. Let's test it with a hundred customers, a hundred calls a day. Let's see how it goes. And so we ran a pilot for a couple of months and just it was able, it made it so we were faster to move on procurement and everything else. And based on that pilot, we were able to grab the actual data of this is how many tokens we consume. To how many calls that we deflect. And then initially we weren't deflecting the best because we hadn't improved and refined it. And so during that couple of months, we were able to improve, refine, improve, refine on a daily basis until we hit higher deflection rates. We were actually resolving calls in about the fifth, a fifth of the time that it could take like a normal live agent to resolve, which was great for our customers. And then we were able to then From there, make informed assumptions to say, okay, if we continue improving at the rate that we're at, then we'll provide this much value, and here's the ROI behind it. At which point we we definitely realize those values and we're continuing to improve it and get more values out of that, which is exciting. Yoni Leitersdorf: That's amazing. Were you were you measuring CSATs or like comparing, you know, how customers were responding to this service versus the previous service? Was that something you you were able to measure? Nate Christiansen: Yes, for sure. That's I mean, that's at the core of it, right? The two things we're saying is, well, can we resolve their issues and do it faster? But the second thing, and probably more importantly, can we provide the customers with the same experience or better experience? And the answer to that is we actually have we provide them 24-7 support, which we didn't before. We provide them multilingual support. Yoni Leitersdorf: Yeah. Yeah. Yeah. Yeah. Nate Christiansen: And they're getting answers to their problem in about a fifth of the time. So, you know, my wife, she's always she she hates to call the hotel, she hates to call Delta Airlines and have me meet she so she has me do it basically. So when I get on there, I care about getting my problem resolved as fast as possible. I Yoni Leitersdorf: Mm-hmm. Mm-hmm. Nate Christiansen: I mean, I do care about talking to a person, but end of the day, like I just want that value to be realized as quickly as possible. And so as we started, we did. We had a lot of customers immediately go, I need to talk to representative. Because we had

a trust issue with Siri and these other agents that they've talked to in phone trees. We changed our intro messages. We've gotten our customers more accustomed to it. And our CSAT this last year is higher than we've ever had it. History of this company. So we are providing our customers with a better support experience than they've ever had before because they're getting their problems resolved faster. They still go to those live technical representatives when there's something very complex, but when there's not and there's just questions, they're getting them answered right away. Yoni Leitersdorf: It's amazing. It's amazing. I I I don't know, but I estimate that even for the customer support reps, it's more interesting to handle, you know, more interesting cases. You know, versus like, you know, how do I do this? Well, you know, on the right, there's a button. You see that it says edit. Yeah. Click on that, right? Like, you know, versus I'm trying to do this, you know, that we I don't know, big load on the system or something complicated. Feels like the humans enjoy get get to work on more interesting and challenging tickets. Does does that sound right? Is that is that yeah. Nate Christiansen: Absolutely. And and to be honest, they are the experts that are saying, you're the AI agent wasn't able to resolve some of those questions. So they're listening and auditing and then they're training it now. so you know, their jobs are molding into being the higher ticket, higher level concierge support and into training the AI, training the model, improving it and Yoni Leitersdorf: Yeah. Nate Christiansen: I'd just say overall they're having a a a great experience with that. Yoni Leitersdorf: It's amazing. And I I think one thing that's that's also kind of stands out, I I I was looking at Henry Shine one's stats before the call. And you know, you don't see like, hey, AI replace all our employees, and then you know, it's you actually see a healthy business that's continuously growing. I guess people get to work on more interesting challenges or maybe you know moving higher up the stack. And so AI is taking The simpler tasks and automating them and giving better service to the customers. And then also your employees get to work on more interesting things. And of course, that drives up motivation. So that's that's amazing to see. So, okay. So there is a financial model, there is ROI, best CSAT ever. what what did you, when you were embarking that journey, What did you expect or think is going to happen and didn't pan out? Like something that you're like, Okay, this is this this will surely work and then it didn't? or some kind of assumption that and we know what they say about assumptions. So anything like that that you can share? Nate Christiansen: Yeah, I think as we as we initially started, I assume that more people would engage with the AI. and now the more I think about it, I'm like, yeah, like My neighbor who's, you know, nearing the end of his career,

who works in dentistry, probably isn't going to be as comfortable the first time he goes and talks to an AI agent. And what was interesting is we found that it was actually the humans that sounded more robotic on those calls often than it was the AI agent. So you would say, Hi, I'm Claire. How can I help you solve your issue? Tell me in a complete sentence. Like, let's get it squared away for you. And you'd have the human say, or the the doctor or whoever it is, say, forms. Tell me more about forms. What's going on with forms? Need help with forms. It's broken. Okay, well, tell me more about your issue. And so, you know, as as powerful as the technology is, this is actually more of a psychology concern. Yoni Leitersdorf: Yeah. Nate Christiansen: That we've been having to address with our customers. And so it's been a combination of marketing and explaining to our customers, them having a few more interactions, and then also just training the AI to like prod them a little bit. before, you know, so one thing that, for example, we're we're looking towards doing, we don't have this yet today, but I would say the hold time, Yoni Leitersdorf: Mm-hmm. Nate Christiansen: If they want to escalate to a live representative, hey, the whole time is three minutes. We'll put you in the queue right now for that hold for those three minutes. But let's just chat for these next three minutes, and then I'll send all that information over to the live rep. And so that's like a great kind of approach because now the customer says, okay, I know I'm getting to a live rep, but I'm on hold. And whatever I share here will be transferred over to that live rep. And as we've done more of the those like live transfer of information into the rep, the customers are gaining more trust that, hey, like this isn't a stupid agent tree phone, phone tree that I've talked to before. Like this is something different. And I was actually visiting an office down in Chicago a couple months ago and they were like were you involved in implementing that AI tool and I was like yeah I was and they're like that's amazing like it's it's so cool like we want it for our practice to answer like our customer questions and concerns so yeah Yoni Leitersdorf: That's amazing. That's amazing. I you know, I've been so many times in the case with like phone trees, and for those listeners who don't know, it's like, you know, when you press this, press five, press seven, you know, in put put your whatever number. And so many times you go through that tree and at the end of it shows up a rep who knows zero about what you're trying to do. Nate Christiansen: Yep. Yoni Leitersdorf: and has not validated you, doesn't know your name, nothing like, okay, what is your phone number? I'm calling from my phone, you know. I built out all this stuff. So having the AI collected and shared with the rep is amazing. you know, surely reduces frustration. Okay. anything else that you know, you may have expected something and it ended up panning out differently on this project before you move to the the the other one? Nate Christiansen: Yeah. Yeah, you know, I I think the biggest battle was not necessarily training the AI and we we had to build the business case, but it's also there's a lot of fear around job security, around you know, the future of my role, if I have tools like this. And, you know, at first I think that was Yoni Leitersdorf: Mm-hmm. Nate Christiansen: not necessarily internal to our company 'cause we have great people, but there was a resistance to like this won't work, this seems risky, this seems scary, what about runaway costs? And so I would just say like Yoni Leitersdorf: And then Nate Christiansen: Having to have that pilot and really prove the financials behind it. And then also just ensuring people with their roles and how it evolves doesn't necessarily go away, it evolves, then got them on the side, right? We we have an incredible support group here at the company and they've done an incredible job and are really some of the key people implementing these tools today. but you definitely have to to make sure. that you align people and don't drive the fear and how to handle and manage that. So. Yoni Leitersdorf: Yeah, yeah, for sure there are I'm sure there are people out there that are concerned about AI, but then and again, throughout a lot of my episodes here, what I discover is that AI actually ends up creating more work, more jobs just elsewhere or higher level. but it it definitely ends up being to to the benefit even of the the people doing the the job previously. Nate Christiansen: Absolutely. Yoni Leitersdorf: Okay, so let's switch over to a second project. So that was on inbound. People calling in, having issues with the product or figuring out how to use a feature, calling in, much better service, they're happier. Now let's look at the other side. So the other side is the AI SDR, which basically means reaching out to potential customers. SDR sales development rep. We have those two at solid. Nate Christiansen: Mm-hmm. Yoni Leitersdorf: They reach out to potential customers. And there's a lot of AI SDRs out there today that are doing it over the phone, email, LinkedIn, whatever, that are able to identify a potential person to sell to. And then they have a story or they craft a story for the person and then either call them or or email or whatever it may be. So in your specific case, I guess first of all, what was the impetus? Was it just, you know, straight trying to sell more? product and get to more customers or was there something else behind there? Nate Christiansen: Yeah, so I think There's honestly, we we did this initiative with support and it's opened a whole new world to us, right? Of opportunities beyond even just SDR, we've got customer success working on it, got training and onboarding, and there's really a lot of different ways that we can support our customers and or sell to them through these this kind of tech. So the the

thought process really came down to, you know, if we're taking over a million calls annually supporting our customers, imagine if you could train your BPO or your support team to turn even a small percentage of all of those support calls into a sales opportunity. It'd be massive, right? And we aren't quite at that point that we've we're We want to keep our support calls in the support world, obviously, but there's a world where, hey, if a customer asks a question about, hey, you have this product, I'm a little bit interested in the support call, that it can create that opportunity and engage some of that conversation. So then what we've done is we have we have a fairly large actually inbound SDR team. So they're taking the inbound calls today of customers interested in purchasing. our product and a lot of their work is to go answer the questions for the customer about their you know product that they're looking into and then to qualify the lead send it set them up with an account executive and that's it however what's hard is we have we have over 60 products here at our company and these people can't necessarily be experts at knowing everything. And so what we're working on right now is we've implemented and trained it on all of our products, but also in our sales motion to understand the customer need, to collect and qualify the lead, to schedule with the account executive. And then now what we're doing is we're Be turning that inbound SDR team into an outbound team that's doing the outbound calls. So we haven't quite begun calling cold calling with an AI to our customers, but we'll take those inbound calls as well as the support calls that turn into sales opportunities and qualify the leads there. Yoni Leitersdorf: Mm-hmm. It's amazing. Again, a another opportunity where you're taking someone who's doing some work and then you are upgrading their work and you know removing because Nate Christiansen: That's right. We're essentially doubling our outbound SDR team by doing that, right? And that's significantly could increase our revenue. Yoni Leitersdorf: Yeah. So we're we're early on, so we don't yet have a measurement of impact of that, right? Of like h how that how that worked. Amazing. Okay, so Yeah. Did you s did you start the pilot of that or not yet? Nate Christiansen: In the process right now. Yeah, we have it piloted. So we we had our last month we had our first closed one opportunity and sale as a result of that. So we we've seen it succeed and we're just being really cautious about our customers and our sales experience because we we obviously don't want customers wanting to buy something and run into an issue or not buy it because they're interacting with an AI. So we're being I would say a little bit more metered and cautious as we roll this one up. Yoni Leitersdorf: Yeah, of course. You wanna you mu you wanna manage that. okay, let's assume this is successful, this project. what's next on the list? Do you have your your your next ideas or you know what where else you would wanna try to apply it? Nate Christiansen: it it's never ending. So I I think one thing that would be interesting is from like a sales demo as well as as a training aspect to do a voice AI and have it also trained on your product. So now instead of having to schedule a sales demo with a person at a very specific time and gather the whole group together, you could have a sales demo at Yoni Leitersdorf: Mm-hmm. Nate Christiansen: middle of the night or on your own whenever you want and the AI that's an expert at the tool will walk you along on how to use a product. so you can imagine that's helpful from a sales side but also from a a training and onboarding or even a support side. Yoni Leitersdorf: Yeah. There's also I've seen the other side of it where you have these tools that mimic the customer. And so it trains your people. And, you know, as a customer, like, okay, I'm a dentist or I'm, you know, I'm part of a network and I'm evaluating solutions. And then and the AI acts as the potential customer and asks your reps the questions and sees how they respond and then scores it after. So they're like the two sides of that interaction that could be AI-ified. Nate Christiansen: Yeah. Yeah, for sure. That'd be awesome. Yoni Leitersdorf: okay. Now it's 2026. You you embarked on the this journey about two years ago. what would you if you had a time machine and you could go and meet Nate two years ago? you know, other than maybe telling him give it like being like you know, back to the future too, where you're like giving him the farmer's almanac. what would you tell him? Nate Christiansen: Mm-hmm. Yoni Leitersdorf: About these projects, what would you tell them to be aware of or do differently or be careful or double down on? What what would you say? Nate Christiansen: Yeah, you know, I think a a lot of people just assume, including myself initially, but just assume that, hey, I'm gonna have this AI tool, I'm gonna plug it in, turn it on, and magically it will just do the work for me. And the reality I've learned is it's more like onboarding an employee, a brand new employee. And so you've gotta be very specific. about what you tell it to do. And if you tell it to do something, it's kind of like the, you know, the genie curse when you you make a wish and then you don't realize that there's consequences to the wish that you just made. sometimes we blame it as a hallucination when in reality Yoni Leitersdorf: Yeah. Yeah. Yep. Nate Christiansen: We just weren't clear on our training. And I would say a new employee, when you're hiring them, they have to make a lot of assumptions and kind of learn their role by talking to different people. And the AI, we don't always have that luxury. And so I think it's very if I was to go back, I would say like Nate, you really need to define Yoni Leitersdorf: Mm-hmm. Nate Christiansen: Everything that you want it to do and make sure that the teams are very robust and clear in the

definitions and then the training that you give it. Otherwise, I mean it would have just sped us up faster because we we had a lot of A B testing where we made mistakes, to be honest. And we're still learning and making mistakes. So that's not bad. I don't regret it. but I don't think AI tooling is always as magical as people assume. There's actually a lot of work that Yoni Leitersdorf: Yeah. I'm smiling. Well, for two reasons. One, first of all, I realized it wasn't the farmer's almanac, it was the sports almanac. So I need to feel I I feel I need to correct myself on this recorded episode. But secondly, we had a a call with a customer today, a solid customer, and they said, you know, we were hoping for it to all to be magical. And the reality is only some of it is magical and a lot of it is still work. Nate Christiansen: Yeah. Yeah. Yoni Leitersdorf: And we said, Yeah, you know, that's that's the case and it's it's it's sort of this is how it works. And you know, everyone's been selling this magic the this magic beans that can do whatever and you can now, you know, grow something and and climb on it. But the reality is that it's it's more complicated than that. for sure. Okay, so that's one thing you would tell yourself. Anything else you would share with yourself? Nate Christiansen: Yeah. Agree. You know, I think I would have jumped a little bit more on on at the time I was being impatient and just wanted to go full throttle, but I realized there's a change management aspect with the people as a part of these tools. And so y you know, a big thing for me going back would be like spend more time not just on implementing the AI and making sure the tech works and the business case is all hashed out, but spend more time with each of the team members that this affects, understand their roles, understand what they do today, and then empathize with them and spend time to like really really understand them. and understand and explain to them how their roles will evolve as a result of this. If you do that, those people, and I hate it that we talk replacing jobs, because that's not the reality, those people are the people that are key to the success of any implementation in these tools that you have. And so the more you talk with them, you gotta go to the very ground of the wherever it's at, the more you talk with them and spend time with them, the more that you're going to train the AI better. You're gonna have their participation and collaboration, and you're gonna have their buy-in, so they're not fighting against you at all. Yoni Leitersdorf: Mm-hmm. Yeah. That's the the buy on is really, really important. and the involvement. Okay. So we're as we're nearing towards the the end of the episode, there's something I like to do. It's a little lightning round. I'm gonna ask you a few questions. Each question you need to answer as fast as you can. You know, the first thing that comes to mind. And then and then explain why why did you pick that answer? Okay. So let's start with this. Nate Christiansen: Yeah. Yoni Leitersdorf: Easy maybe. AI tool you can't live without. Nate Christiansen: So I just vibe coded a little tool. We we believe in sustainability at my house. So I I vibe coded a little tool that helps me take the photos from my camera roll and separate some of those items and then repost them onto some of these like Poshmark or these other sites that you can sustainably resell some of your clothing or items. And it saved my wife a lot of time when she goes through the closet because she can immediately separate all the items and not have to upload them. It just kinda auto does it all for her. And so I guess the answer to that is using Claude Code, cursor, whatever it is. I've used Kiro on through AWS, but just being able to build my own little tools, I've never felt more creative and empowered to do that. Yoni Leitersdorf: That's amazing. we should talk about it after because my my wife has a project too where she she wants to build this AI that will tell you what to wear. Like you can say, I'm going to this event, and it would know your inventory. You, you know, you first need to do like your wardrobe inventory, but then it will say, you know, hey, Yunny, based on what I know of you and how you look like, etc., skin tone, whatever. And then, you know, you want to go to this event, this is what I recommend. and you know, accessories and everything. So right now it's my ten year old. Nate Christiansen: Right. That's awesome. Yoni Leitersdorf: Vibe coding that, so we'll see. We'll see how that goes. okay, an AI myth you'd bust in one sentence. Nate Christiansen: Yeah, I think the myth of all the jobs are gonna be replaced with AI tools. I I think just I I was actually trying to explain to my daughter like relativity and quantum physics, because I I did mechanical engineering before. And what's crazy is back in the nineteen hundreds when those were first invented or discovered, Computers didn't exist as we know it today. But because of quantum physics, we have AI. We have all these technology. We have all these tools. And so I believe with AI, we're gonna make new discoveries as humanity that's going to be that next computer, that next technology, healthcare, whatever it is. Like this is going to improve the world and and create more jobs than we've ever seen before. I really believe that. Yoni Leitersdorf: Thank you. I agree. one prediction for AI in the next five years, except for what you just said. Nate Christiansen: one. Your question there. Yeah, I I think we're gonna have another chat GBT type of moment when you know everyone was like, Whoa, this is amazing, this is life-changing. But I think it's gonna be with spatial and physical robotics and AI. I think there's a whole nother world. We're looking at it in manufacturing and Yoni Leitersdorf: Yeah, exactly. So now we need to Nate Christiansen: Car manufacturing is a big area where that's already kind of in development. But I think

our world is going to change and develop rapidly when it comes to like the actual physical benefits of having AI tools. Yoni Leitersdorf: There is I don't know if this robot is still doing this, but have you heard of the figure AI robot? Have you heard of that thing? Yeah, it's just flipping packages all day. And I don't know if it's like it was supposed to start a thing for a few hours, and then they're like, this works, and the robot's not quitting. So keep doing it. And and I think I was watching it three or four days later, and the robot was still. Nate Christiansen: Yes. Yeah, it's crazy. Ha ha. Yoni Leitersdorf: Flipping packages. and and just you know, it was it was a cool thing to watch. Yes, it's very early. Yes, it's you know, just there's so much more to go, but I think I agree with you. The potential there is huge. and a few years from now, maybe robots instead of flipping packages, they'll do other things for us. and that would be amazing. okay. So we're at the end of our episode. This was really enjoyable and was also great to hear. Nate Christiansen: Mm-hmm. Yeah. Yoni Leitersdorf: the success and I'm and I'm very happy that you you guys were able to take an AI pilot, get into production, see the value from it, see the impact on the metrics. Also, employees are enjoying that a lot better. Seems like a great win. Employees and customers and everyone, great win all around. so Nate, I really appreciate you getting on the podcast today. and then Anyone who's listening to this podcast or viewing it on YouTube, please remember to like, subscribe, five-star review, whatever it is on the platform you're listening. and then join us again in our next episode. Thank you very much. Thanks, Nate. Have a good day. Thanks, bye-bye. Nate Christiansen: Thank you. Appreciate it. Yoni Leitersdorf: call the Henry Shine One support line and see how the experience was. So let's give it a try. Thank you for calling Henry Shine One Support. Your call may be monitored or recorded for quality or training purposes. Please listen closely as the menu options have changed. Press one for Dentrix, e-services, or Dentrix Enterprise. Press one for Dentrix or e-services. Press two. Thank you for calling HenryShine One Support for assistance with claim processing. I'll make up one. going on and let's get that resolved for you. my name is Yoni and I'm trying to log in. It's nice to meet you, Yoni. Which product are you trying to log into? good question. Which ones are there that I could log into? We offer several products like Dentrix, Ascend, Enterprise, and Seth Me. Are you using one of those or perhaps a patient communication tool like Patient Engage or Lighthouse? Dentrix Ascend. Got it. Let me take a closer look at that for you. Since you're using Dentrix, are you trying to log into the software on your computer or are you trying to access the online resource center? trying to log on my computer. To log in on your computer, start by double clicking a Dentrix icon on your desktop, like the Office Manager. Are you seeing a login box asking for a user ID and password? Nope, I double clicked and I saw an empty screen. That sounds like the module might be opening off screen or is stuck in the background. Let's try to bring it back by pressing the Alt and Spacebar keys at the same time and then select Maximize from the menu that appears. Did that bring the login window into view for you? Yeah, it did. Since that didn't bring it up, let's check if the computer thinks it's already running. Can you press control shift and escape all at the same time to open your task manager? Done. Now what? Great. Now that you have the task manager open, look for the Dentrix module you tried to open under the processes tab. If you see it there, right-click it and select End Task, then try opening the module again from your desktop. Did that do the trick? Yeah, it did. You know, you're amazing. You're really good. I'm so glad that worked for you. Since you're back in, is there anything else I can help you with today? can I get pricing information about the product? I'm kind of curious how much this costs annually. I don't have access to your specific contract or annual pricing details here in support. To get the most accurate information for your practice, I can connect you with a customer success manager who can review your account and pricing with you. Would you like me to get someone on the line for you? No, you've helped me enough. Have a good day. one question. Are you an AI? I am. I'm an AI phone support agent for Henry Shine One. It's been a pleasure helping you get back into your dentrix software today. Before we finish, I'd love to get your feedback. On a scale of one to five, where five is very satisfied, how would you rate your experience with me today? Please say your rating. Five. Thanks for your response. Have a great day. Goodbye for now. That's the future, guys.