WEBVTT 00:00:02.350 --> 00:00:05.070 Welcome to the Growing EBITDA Podcast, where 00:00:05.070 --> 00:00:07.669 we unlock the doors to management and technology 00:00:07.669 --> 00:00:10.689 insights in the middle market. Join us as we 00:00:10.689 --> 00:00:13.230 explore innovative strategies to drive revenue 00:00:13.230 --> 00:00:16.129 and EBITDA growth, interviewing industry leaders 00:00:16.129 --> 00:00:18.949 and technology experts. Whether you're looking 00:00:18.949 --> 00:00:21.809 to streamline operations, understand the latest 00:00:21.809 --> 00:00:24.629 tech trends, or lead your company towards exponential 00:00:24.629 --> 00:00:28.050 growth, you're in the right place. Stay tuned 00:00:28.050 --> 00:00:34.079 and let's grow together. So James, let's start 00:00:34.079 --> 00:00:36.939 with a question. Yeah. When are we going to get 00:00:36.939 --> 00:00:39.719 some leaders and some technology experts on this 00:00:39.719 --> 00:00:42.359 cast? I know, I was just listening to the intro. 00:00:42.640 --> 00:00:44.399 People are waiting. I know, it's embarrassing. 00:00:44.579 --> 00:00:47.039 I'm like, oh no, we made a lot of promises on 00:00:47.039 --> 00:00:50.079 this thing and they get stuck with us. That's 00:00:50.079 --> 00:00:53.200 a good point. A lot of hollow promises. We'll 00:00:53.200 --> 00:00:55.219 go back and edit that one. A lot of empty ones. 00:00:56.179 --> 00:00:58.759 We'll go ahead with that. Yeah, yeah. So what 00:00:58.759 --> 00:01:01.259 are we talking about today, James? Yeah, you 00:01:01.259 --> 00:01:04.340 know, so right now, Mike, I'm calling in, I guess, 00:01:04.379 --> 00:01:07.500 potting in from Nashville, visiting some clients. 00:01:07.859 --> 00:01:10.000 And I just had a conversation today with a client. 00:01:10.019 --> 00:01:12.340 So I thought maybe we could talk about AI today 00:01:12.340 --> 00:01:14.780 since it seems like it's something everyone's 00:01:14.780 --> 00:01:16.219 trying to figure out. And more and more as we 00:01:16.219 --> 00:01:18.120 go and meet with clients, we get asked about 00:01:18.120 --> 00:01:20.040 it. And since it's kind of on the top of my mind, 00:01:20.120 --> 00:01:21.879 if you're okay with it, I thought maybe we could 00:01:21.879 --> 00:01:24.239 talk about it a little bit. Let's do it. Let's 00:01:24.239 --> 00:01:26.159 do it. Well, yeah, it is certainly everywhere. 00:01:26.200 --> 00:01:29.769 I feel like probably because of... Knowing you, 00:01:29.870 --> 00:01:32.870 I was an early adopter of ChatGPT. That was really 00:01:32.870 --> 00:01:37.069 kind of my first AI exposure. And I have to be 00:01:37.069 --> 00:01:41.030 one of their top evangelists and highest users. 00:01:41.730 --> 00:01:43.689 I'm not going to say that it does my job, but 00:01:43.689 --> 00:01:45.510 it certainly makes it a lot easier these days. 00:01:45.890 --> 00:01:48.969 And I can remember, you know, you go back 18 00:01:48.969 --> 00:01:53.030 months, 12 months ago, probably now. And I would 00:01:53.030 --> 00:01:55.129 be talking about it at board meetings or with 00:01:55.129 --> 00:01:58.109 other executives and just singing its praises. 00:01:58.750 --> 00:02:00.549 candidly was getting a lot of blank stares. People 00:02:00.549 --> 00:02:01.989 had heard of it. It was in the media every day, 00:02:02.069 --> 00:02:05.129 but I would say nine out of 10 executives that 00:02:05.129 --> 00:02:07.609 I was speaking with 18 months ago, weren't using 00:02:07.609 --> 00:02:11.349 it at all and had never used it. And we're cautiously 00:02:11.349 --> 00:02:13.629 interested in maybe learning a little bit, a 00:02:13.629 --> 00:02:16.150 little bit more fast forward to today. I was 00:02:16.150 --> 00:02:18.509 in a board meeting this week and we spent about 00:02:18.509 --> 00:02:20.930 an hour talking about AI and how that's going 00:02:20.930 --> 00:02:23.189 to impact the business. Same thing last week 00:02:23.189 --> 00:02:25.610 in another, in another session. So I think, you 00:02:25.610 --> 00:02:29.099 know, certainly the The popularity is there now. 00:02:29.419 --> 00:02:31.699 Dare I say adoption yet? I think a lot of people 00:02:31.699 --> 00:02:34.759 who maybe hadn't used it a year ago now are kind 00:02:34.759 --> 00:02:37.300 of just finding their sea legs, so to speak, 00:02:37.379 --> 00:02:40.360 starting to dabble with it, including folks who 00:02:40.360 --> 00:02:43.139 I know that I know them not to be technologists, 00:02:43.319 --> 00:02:45.659 right? So a lot of kind of technology lay people, 00:02:45.819 --> 00:02:48.520 if you will, are starting to use it a lot more. 00:02:49.020 --> 00:02:50.800 And did a little bit of research just to kind 00:02:50.800 --> 00:02:53.060 of bait the hook for us for today's conversation. 00:02:54.409 --> 00:02:57.030 It looks like about 70 plus percent of mid -market 00:02:57.030 --> 00:03:01.270 executives are saying that they plan a significant 00:03:01.270 --> 00:03:04.789 increase in AI investments in 2025. So not really 00:03:04.789 --> 00:03:06.189 sure what that means. I'm sure it means a lot 00:03:06.189 --> 00:03:07.389 of different things to a lot of different people. 00:03:07.729 --> 00:03:09.270 I think one of the challenges is going to be, 00:03:09.310 --> 00:03:13.250 how do you show ROI, right? My opinion, you have 00:03:13.250 --> 00:03:16.090 tools like ChatGPT, ROI is off the charts. I 00:03:16.090 --> 00:03:18.930 pay for whatever the expensive version is, like 00:03:18.930 --> 00:03:22.379 20 bucks a month. huge roi but certainly and 00:03:22.379 --> 00:03:24.039 we'll get into this a little bit later we're 00:03:24.039 --> 00:03:26.180 seeing a lot of ai tools out there that are looking 00:03:26.180 --> 00:03:29.180 like cadillacs that can be very offensive so 00:03:29.180 --> 00:03:31.319 be interesting to see kind of where people are 00:03:31.319 --> 00:03:33.919 finding bang for their buck as a whole new industry 00:03:33.919 --> 00:03:37.500 springs up so excited to talk about it and we'll 00:03:37.500 --> 00:03:38.860 dive into it a little bit here well i like i 00:03:38.860 --> 00:03:41.259 like stats like that though the 70 plan to increase 00:03:41.259 --> 00:03:44.199 because uh As a guy who just recently had a child, 00:03:44.360 --> 00:03:47.099 let me tell you, having a baby, I now spend more 00:03:47.099 --> 00:03:48.780 on babies than I've ever spent in my life. And 00:03:48.780 --> 00:03:51.500 so I think when AI didn't exist before, when 00:03:51.500 --> 00:03:53.560 you have stats like you're 70 % now planning 00:03:53.560 --> 00:03:55.560 to spend more, well, it's like, well, to your 00:03:55.560 --> 00:03:57.280 point, one out of 10 was probably spending something 00:03:57.280 --> 00:03:59.580 last year. Now we see that that brought forward 00:03:59.580 --> 00:04:02.300 a bit. And I think that the interesting thing 00:04:02.300 --> 00:04:05.590 when we have these conversations is... Not to 00:04:05.590 --> 00:04:08.229 date us, Mike, but a little bit of time. I can 00:04:08.229 --> 00:04:09.949 remember when the internet came into fruition 00:04:09.949 --> 00:04:12.969 and like I was using the Apple IIe playing Oregon 00:04:12.969 --> 00:04:16.529 Trail at school. And then like this crazy thing, 00:04:16.629 --> 00:04:19.670 Yahoo, came up and I was Googling, I guess Yahooing 00:04:19.670 --> 00:04:21.810 information and learning things from the web. 00:04:21.949 --> 00:04:23.829 And I kind of think through our time, there's 00:04:23.829 --> 00:04:26.449 been a ton of transformation. Smartphone, tablet. 00:04:27.129 --> 00:04:29.329 We know all the things, but that kind of like 00:04:29.329 --> 00:04:32.170 true core, what's really changing business a 00:04:32.170 --> 00:04:34.509 bit. It's interesting. This is the big one that 00:04:34.509 --> 00:04:36.290 I remember because we went through blockchain. 00:04:36.389 --> 00:04:37.949 We went through machine learning and machine 00:04:37.949 --> 00:04:39.730 learning still around to an extent, but went 00:04:39.730 --> 00:04:41.769 through these kind of more trend things. And, 00:04:41.769 --> 00:04:44.829 you know, if we remember like EFTs and all that, 00:04:44.889 --> 00:04:47.149 now we're into this point where we're like, hey, 00:04:47.209 --> 00:04:48.889 this we think is really going to hang around. 00:04:49.600 --> 00:04:52.040 And it's this AI. So I think finding a way that 00:04:52.040 --> 00:04:54.079 really doesn't the buzz side that's truly tied 00:04:54.079 --> 00:04:56.839 to operations is a great place and why I'm excited 00:04:56.839 --> 00:04:59.439 to get it today. So maybe you can kind of tee 00:04:59.439 --> 00:05:02.740 us up for our first topic. Yeah. So topic number 00:05:02.740 --> 00:05:06.399 one, where are we seeing it add real value, right? 00:05:06.500 --> 00:05:09.040 We're still early days. A lot of folks are going 00:05:09.040 --> 00:05:11.560 to figure this out, try a lot of different options 00:05:11.560 --> 00:05:13.019 out there in the coming years, I'm sure. But 00:05:13.019 --> 00:05:15.620 where are you actually seeing it move the needle 00:05:15.620 --> 00:05:19.060 today in mid -market operating companies? Yeah. 00:05:19.720 --> 00:05:21.579 To go back a little bit, and you said you sit 00:05:21.579 --> 00:05:23.079 in boards, which I think is a lot of the times 00:05:23.079 --> 00:05:24.759 we have conversation around what you're seeing 00:05:24.759 --> 00:05:26.439 and then how we're supporting other boards and 00:05:26.439 --> 00:05:28.699 allows you and I to kind of have a dialogue of 00:05:28.699 --> 00:05:30.660 what's going on out there in the business space. 00:05:31.139 --> 00:05:33.540 We started a program because we were asked about 00:05:33.540 --> 00:05:36.879 it so often by PE firms to do these one -day 00:05:36.879 --> 00:05:39.399 workshops where we go through the business needs 00:05:39.399 --> 00:05:41.540 and how AI can be a part of those business needs 00:05:41.540 --> 00:05:44.199 and help folks out. And so part of this program 00:05:44.199 --> 00:05:46.360 is we go through and explain what we've seen 00:05:46.360 --> 00:05:48.120 in the market and we've seen it be successful. 00:05:49.040 --> 00:05:50.639 it's good to go through a few examples that have 00:05:50.639 --> 00:05:53.040 come out of those discovery sessions and some 00:05:53.040 --> 00:05:55.160 tools that we've used that if folks want to know 00:05:55.160 --> 00:05:57.220 what the tool is, feel free to reach out to us. 00:05:57.300 --> 00:05:58.740 But we're going to speak kind of generally today 00:05:58.740 --> 00:06:01.819 about the tool sets that we think align to the 00:06:01.819 --> 00:06:03.920 different options. But kind of, I would say, 00:06:04.040 --> 00:06:06.759 three, maybe four buckets we felt that folks 00:06:06.759 --> 00:06:10.100 have been pretty successful. Number one, we've 00:06:10.100 --> 00:06:13.060 done a lot of work with helping companies take 00:06:13.060 --> 00:06:16.339 large amounts of data and distill that data down 00:06:16.339 --> 00:06:18.439 for simple consumption. So where we see that 00:06:18.439 --> 00:06:20.779 a lot is in customer service. I have a customer 00:06:20.779 --> 00:06:22.759 service rep who needs to understand all their 00:06:22.759 --> 00:06:25.139 products, how to support, what needs to be done, 00:06:25.220 --> 00:06:27.839 and they need a simple answer. Putting a chatbot 00:06:27.839 --> 00:06:29.860 on top of a large amount of data where I can 00:06:29.860 --> 00:06:32.959 ask the chatbot natural questions like, when 00:06:32.959 --> 00:06:35.120 an order comes back damaged, what should I do? 00:06:35.279 --> 00:06:37.319 And have it very simply respond to help the client. 00:06:37.480 --> 00:06:39.839 We've seen that be successful. And that call 00:06:39.839 --> 00:06:42.180 and response is a little bit on top of what's 00:06:42.180 --> 00:06:44.439 already in the market before AI, but now much 00:06:44.439 --> 00:06:48.509 richer and much more efficient. Number two. is 00:06:48.509 --> 00:06:51.029 around sales and inventory planning. So we've 00:06:51.029 --> 00:06:53.730 all seen those kind of like algorithms in the 00:06:53.730 --> 00:06:56.189 past, I would say seasonality, ship times, lead 00:06:56.189 --> 00:06:58.449 times, and help folks plan their procurement 00:06:58.449 --> 00:07:01.129 side. Well, with some of these AI tools, we can 00:07:01.129 --> 00:07:03.870 now start to see, well, it's a holiday around 00:07:03.870 --> 00:07:06.449 this time, so things go slower. The Panama Canal 00:07:06.449 --> 00:07:08.449 typically goes under maintenance at this point, 00:07:08.490 --> 00:07:10.089 so that container is going to be slower. And 00:07:10.089 --> 00:07:11.930 it really starts to bring some more rich data 00:07:11.930 --> 00:07:14.790 to help folks plan better and have more detailed 00:07:14.790 --> 00:07:16.810 assumptions when they're trying to plan their 00:07:16.810 --> 00:07:19.189 inventory and procurement. procurement. A third 00:07:19.189 --> 00:07:23.089 one is around finance. A lot of times people 00:07:23.089 --> 00:07:26.350 have used OCR, which allows you to receive documents, 00:07:26.610 --> 00:07:28.870 scan those documents, and pull the text out of 00:07:28.870 --> 00:07:32.350 those documents. The OCR process got that in 00:07:32.350 --> 00:07:34.230 the system pretty quickly, but you still had 00:07:34.230 --> 00:07:36.529 to take a manual process to match that document 00:07:36.529 --> 00:07:38.829 to the right provider and make sure that the 00:07:38.829 --> 00:07:41.470 information was close. Now, with some of the 00:07:41.470 --> 00:07:43.790 AI tools, you can quickly load that document. 00:07:43.850 --> 00:07:46.069 It will serve up an answer that assumes to be 00:07:46.069 --> 00:07:48.709 correct. A human comes in and touches it, makes 00:07:48.709 --> 00:07:50.509 sure it is, and then it moves on in the process. 00:07:50.910 --> 00:07:52.990 And so I think those, when you think about those, 00:07:53.050 --> 00:07:55.230 almost like what we see in the warehouse, like 00:07:55.230 --> 00:07:57.810 cobots, where AI is bringing information to the 00:07:57.810 --> 00:08:00.790 surface and a human's taking action, is really 00:08:00.790 --> 00:08:03.560 where we see that actually move the needle. I 00:08:03.560 --> 00:08:06.759 dream of fully automated where AI is doing it 00:08:06.759 --> 00:08:08.639 from the initial question till it's all the way 00:08:08.639 --> 00:08:11.060 answered. Probably isn't a reality. There's not 00:08:11.060 --> 00:08:12.959 a tool that does that. But we have seen some 00:08:12.959 --> 00:08:14.720 clients be very successful with getting those 00:08:14.720 --> 00:08:16.920 answers brought to humans so humans can act. 00:08:17.879 --> 00:08:22.220 Yeah, and I think what I'm seeing is the exact 00:08:22.220 --> 00:08:25.100 same thing play out more at the executive level. 00:08:25.339 --> 00:08:27.420 I think we're seeing even some of these simple 00:08:27.420 --> 00:08:30.920 tools like ChatGPT, I'm seeing a lot of executives 00:08:30.920 --> 00:08:34.370 being able to dump in business strategies. Even 00:08:34.370 --> 00:08:37.350 a simple tool like ChatGPT, you can load in, 00:08:37.429 --> 00:08:40.610 hey, here's this strategy that we're thinking 00:08:40.610 --> 00:08:42.590 about, or here's this problem that we're wrestling 00:08:42.590 --> 00:08:45.730 with. Obviously, these tools can be phenomenal 00:08:45.730 --> 00:08:48.690 if you prompt them correctly in really helping 00:08:48.690 --> 00:08:51.730 you refine your thought. Loading in a bunch of 00:08:51.730 --> 00:08:54.190 data, a bunch of context, and then challenging 00:08:54.190 --> 00:08:57.149 it to stimulate your thinking or challenge your 00:08:57.149 --> 00:08:59.129 thinking. And certainly that's how I'm using 00:08:59.129 --> 00:09:01.490 it on a daily basis, not just having it write 00:09:01.490 --> 00:09:04.529 95 % of my emails. If I could get to 100, man, 00:09:04.730 --> 00:09:06.990 we'd be cooking with fire. Yeah, you're right. 00:09:07.049 --> 00:09:08.789 And that's that call response a bit, right? And 00:09:08.789 --> 00:09:11.009 that's interesting. You as an individual using 00:09:11.009 --> 00:09:13.330 that, again, as an executive over a mid -market 00:09:13.330 --> 00:09:15.870 company, that's a prime example of a very successful 00:09:15.870 --> 00:09:17.710 use case. One of the things we'll talk about 00:09:17.710 --> 00:09:20.210 later in the podcast is how to do that in a wise 00:09:20.210 --> 00:09:21.590 fashion, which I know you and I have had those 00:09:21.590 --> 00:09:23.529 conversations to make sure you're not loading 00:09:23.529 --> 00:09:26.389 the incorrect data and doing it correctly. But 00:09:26.389 --> 00:09:28.850 I think that's a very practical one. And so maybe 00:09:28.850 --> 00:09:30.850 we go to our next topic, which is since you brought 00:09:30.850 --> 00:09:34.049 up a practical tool like ChatGPT, where is AI 00:09:34.049 --> 00:09:35.909 kind of still a bit of hype in your opinion? 00:09:35.950 --> 00:09:37.570 Because obviously you go to these board meetings 00:09:37.570 --> 00:09:40.049 and there's a lot of truth in a board meeting, 00:09:40.090 --> 00:09:41.450 but there's a little bit of hype and a little 00:09:41.450 --> 00:09:43.190 bit of salespersonship where they're trying to 00:09:43.190 --> 00:09:46.830 kind of... drive you to them or you walking away 00:09:46.830 --> 00:09:48.470 feeling like they know more than they maybe do 00:09:48.470 --> 00:09:50.610 in some spaces. Where do you feel the hype is 00:09:50.610 --> 00:09:52.149 in AI and kind of what have you been able to 00:09:52.149 --> 00:09:54.529 tease out and maybe double click on and understand 00:09:54.529 --> 00:09:57.909 where they've overhyped it? I mean, specifically, 00:09:57.909 --> 00:10:00.409 one of the things that I've come across a lot, 00:10:00.490 --> 00:10:02.289 just some things that I've been working on of 00:10:02.289 --> 00:10:04.529 late, there's a lot of marketing automation, 00:10:04.850 --> 00:10:08.669 AI tools out there, which quite frankly, are 00:10:08.669 --> 00:10:10.909 doing, in my opinion, what ChatGPT already does 00:10:10.909 --> 00:10:13.350 quite well. Right. And they're doing it for an 00:10:13.350 --> 00:10:16.110 extremely expensive number. You may recall this 00:10:16.110 --> 00:10:18.450 conversation I had a few months ago where I was 00:10:18.450 --> 00:10:21.960 helping one of our businesses think about. partnering 00:10:21.960 --> 00:10:24.740 up with a kind of a software vendor to help them 00:10:24.740 --> 00:10:26.899 explore a kind of a new strategy for themselves 00:10:26.899 --> 00:10:29.580 or for one part of their business. And I shared 00:10:29.580 --> 00:10:31.360 the partner with you and you're like, hey, whoa, 00:10:31.440 --> 00:10:34.559 watch out. Anybody with a .ai at the end of their 00:10:34.559 --> 00:10:37.159 name, like scrutinize them heavily, right? Because 00:10:37.159 --> 00:10:40.120 there is just a lot of hype out there. And some 00:10:40.120 --> 00:10:43.419 of the platforms that I've seen do exactly what... 00:10:43.870 --> 00:10:46.529 ChatGPT does for free with a little bit of polish 00:10:46.529 --> 00:10:49.330 on the margin, admittedly, but they're doing 00:10:49.330 --> 00:10:51.789 what ChatGPT does for free or for 20 bucks a 00:10:51.789 --> 00:10:55.590 month for 20, 40, 50, $100 ,000 a year for an 00:10:55.590 --> 00:10:58.210 enterprise grade tool. And quite frankly, I think 00:10:58.210 --> 00:11:00.970 they obviously have clients, they're companies 00:11:00.970 --> 00:11:02.809 that are in business, hopefully they're doing 00:11:02.809 --> 00:11:05.789 well. We want everybody to succeed in this life. 00:11:05.870 --> 00:11:07.529 But I got to be honest with you, I think there 00:11:07.529 --> 00:11:08.929 is a lot of hype around some of these tools. 00:11:09.049 --> 00:11:11.190 I think you got to really ask yourself, could 00:11:11.190 --> 00:11:13.840 somebody take a few hours, which they should 00:11:13.840 --> 00:11:16.340 now have a few hours extra if they're using these 00:11:16.340 --> 00:11:19.240 types of tools productively. Could they take 00:11:19.240 --> 00:11:21.960 a couple of those now unused hours and throw 00:11:21.960 --> 00:11:24.240 some content into ChatGPT and have that spit 00:11:24.240 --> 00:11:27.259 out some really neat marketing stuff versus paying 00:11:27.259 --> 00:11:29.580 for a Cadillac? I think right now the Honda, 00:11:29.840 --> 00:11:33.159 so to speak, is like nine -tenths of the Cadillac. 00:11:33.299 --> 00:11:35.980 And before folks really start shelling out big 00:11:35.980 --> 00:11:38.799 dollars for some of the more expensive and admittedly 00:11:38.799 --> 00:11:40.700 a little bit more elegant solutions, I think 00:11:40.700 --> 00:11:43.600 you really got to kind of canvas the landscape 00:11:43.600 --> 00:11:45.960 and say, Hey, which one of these companies do 00:11:45.960 --> 00:11:47.000 we think is actually going to be in business 00:11:47.000 --> 00:11:49.000 in three or four years? How much do we actually 00:11:49.000 --> 00:11:52.440 want to invest in this? And is there a free or 00:11:52.440 --> 00:11:56.019 near free solution? The open AI tool set goes 00:11:56.019 --> 00:12:00.100 well beyond chat GPT, image processing, transcripting, 00:12:00.100 --> 00:12:02.480 things like that, which if we just stick in here 00:12:02.480 --> 00:12:04.720 in marketing for a second, those are some tools 00:12:04.720 --> 00:12:06.759 that many marketers need. And most of those are 00:12:06.759 --> 00:12:08.779 available for free or near free. You just got 00:12:08.779 --> 00:12:10.840 to configure them up a little bit, which the 00:12:10.840 --> 00:12:12.840 population of folks that can do that is, is growing 00:12:12.840 --> 00:12:15.059 quite rapidly, right? Independence as well as 00:12:15.059 --> 00:12:17.120 consultants. So I think you just gotta, I think 00:12:17.120 --> 00:12:18.740 you just gotta watch out for some of this, for 00:12:18.740 --> 00:12:21.870 some of these hyped solutions out there. Yeah. 00:12:21.970 --> 00:12:23.549 So, Mike, one of the things that I think was 00:12:23.549 --> 00:12:24.769 really great that you've done is you've built 00:12:24.769 --> 00:12:26.710 some workflows for yourself for some repetitive 00:12:26.710 --> 00:12:29.129 tasks that you've had that you said, hey, I'm 00:12:29.129 --> 00:12:32.429 going to go kind of apply this AI tool to help 00:12:32.429 --> 00:12:34.950 me in this very specific way. And we don't have 00:12:34.950 --> 00:12:36.730 to get into all the use cases unless you want 00:12:36.730 --> 00:12:38.570 to. But I think one of the things I'd be curious 00:12:38.570 --> 00:12:40.850 to hear from you, because you kind of self -serve 00:12:40.850 --> 00:12:42.850 this and did a bit of your own training and your 00:12:42.850 --> 00:12:46.149 discovery. So if I'm an executive like yourself 00:12:46.149 --> 00:12:48.330 and I'm like, hey, I got a few repetitive workflows 00:12:48.330 --> 00:12:50.350 that Mike mentioned today that feels like they 00:12:50.350 --> 00:12:51.909 could be automated. And maybe you could just 00:12:51.909 --> 00:12:53.590 take us on a quick little journey of how you 00:12:53.590 --> 00:12:55.450 went through that discovery and how you learned 00:12:55.450 --> 00:12:58.149 to use the tool. And kind of, you know, I think 00:12:58.149 --> 00:12:59.590 one of the things I do know is you do experiment 00:12:59.590 --> 00:13:01.269 and failure, right? Like you tested out, didn't 00:13:01.269 --> 00:13:03.090 work, move on to the next. But like, what are 00:13:03.090 --> 00:13:04.710 some other tactics that maybe you use that you 00:13:04.710 --> 00:13:06.250 could share with our listeners of like, hey, 00:13:06.309 --> 00:13:08.210 get out there, get your hands dirty, try some 00:13:08.210 --> 00:13:10.610 things. And then, you know, maybe align and find 00:13:10.610 --> 00:13:12.409 something that works well like you do. Yeah, 00:13:12.450 --> 00:13:15.090 I mean, I think it doesn't have to be an executive 00:13:15.090 --> 00:13:17.750 task per se. The example you're talking about 00:13:17.750 --> 00:13:20.399 was something that I do all the time. I was looking 00:13:20.399 --> 00:13:23.220 for a kind of elegant, automated, low -cost solution. 00:13:23.700 --> 00:13:27.100 I ended up figuring out how to take a process 00:13:27.100 --> 00:13:30.019 that took me about 20 to 30 minutes manually 00:13:30.019 --> 00:13:31.779 every time that I did it. And I did it, you know, 00:13:31.779 --> 00:13:34.259 four or five times a week. I was able to get 00:13:34.259 --> 00:13:37.480 that down to somewhere in the ballpark of five 00:13:37.480 --> 00:13:40.200 to six seconds per instance, right? So I wake 00:13:40.200 --> 00:13:42.399 up with four or five hours or maybe three or 00:13:42.399 --> 00:13:45.120 four hours, whatever the math is, more time available 00:13:45.120 --> 00:13:47.419 to go and do other things. But I really, I think 00:13:47.419 --> 00:13:50.129 it comes down to... basic fundamental lean principles, 00:13:50.269 --> 00:13:52.809 right? If you think about it from a process optimization 00:13:52.809 --> 00:13:55.590 perspective, if you've got workflows in your 00:13:55.590 --> 00:14:00.120 business where data is involved. there is a high 00:14:00.120 --> 00:14:02.980 likelihood that even a basic tool like ChatGPT, 00:14:03.139 --> 00:14:06.779 maybe in a closed environment or Copilot in a 00:14:06.779 --> 00:14:08.659 closed environment, you can take some of these 00:14:08.659 --> 00:14:11.220 basic tools and really streamline some of these 00:14:11.220 --> 00:14:13.139 workflows. And I think you just got to kind of 00:14:13.139 --> 00:14:15.139 challenge yourself to say, hey, let's look across 00:14:15.139 --> 00:14:17.600 this business function by function and say, hey, 00:14:17.620 --> 00:14:19.659 is there something that that person over there 00:14:19.659 --> 00:14:22.320 is doing regularly that we could just spend a 00:14:22.320 --> 00:14:24.519 few minutes process mapping and maybe we need 00:14:24.519 --> 00:14:26.879 to find a resource to help us stitch some of 00:14:26.879 --> 00:14:29.789 these. discrete software programs that we're 00:14:29.789 --> 00:14:31.970 using as part of that workflow. Maybe it's CRM, 00:14:31.990 --> 00:14:34.610 maybe it's a tool as simple as notes or something 00:14:34.610 --> 00:14:37.169 like that, right? How do we actually stitch these 00:14:37.169 --> 00:14:38.549 things together? There's a bunch of folks out 00:14:38.549 --> 00:14:40.009 there with some neat tools that they're using 00:14:40.009 --> 00:14:42.350 that they can do that and just look for waste, 00:14:42.470 --> 00:14:45.590 right? Look for waste, map out the process, try 00:14:45.590 --> 00:14:48.169 and find an automated AI enabled solution. And 00:14:48.169 --> 00:14:49.990 you can build some of this stuff for literally 00:14:49.990 --> 00:14:53.470 pennies relative to the amount of time that you're 00:14:53.470 --> 00:14:56.429 going to save. So again, it's early days, but 00:14:56.970 --> 00:14:58.710 we're starting to see some of these things emerge 00:14:58.710 --> 00:15:01.690 where even some executives like myself are able 00:15:01.690 --> 00:15:04.490 to cut out some non -value added time and pick 00:15:04.490 --> 00:15:06.269 up some extra time to go and be productive elsewhere. 00:15:06.870 --> 00:15:09.049 As folks probably know, one of the things we 00:15:09.049 --> 00:15:10.850 do is we do a lot of diligence work. And as part 00:15:10.850 --> 00:15:13.029 of our diligence work, we receive a lot of large 00:15:13.029 --> 00:15:15.809 data sets from folks. And so one of the things 00:15:15.809 --> 00:15:16.830 we're going to talk about in a second is around 00:15:16.830 --> 00:15:19.850 that safety, security, privacy side of AI. But 00:15:19.850 --> 00:15:21.629 putting that aside for a second, talk about how 00:15:21.629 --> 00:15:23.190 I have this large data set and I need to work 00:15:23.190 --> 00:15:24.730 through it. I'm not going to put that data in 00:15:24.730 --> 00:15:27.470 AI, but I want to learn a way to better comb 00:15:27.470 --> 00:15:29.230 through that data. I'll tell you, one of the 00:15:29.230 --> 00:15:31.529 number one use cases that we hear when we go 00:15:31.529 --> 00:15:35.049 and have these conversations is PE firms using... 00:15:35.210 --> 00:15:37.850 the ai solution to help them build better excel 00:15:37.850 --> 00:15:40.070 formulas which is interesting right because you're 00:15:40.070 --> 00:15:42.570 still using excel which is a highly manual tool 00:15:42.570 --> 00:15:44.909 you're not using a full integrated system but 00:15:44.909 --> 00:15:47.149 now you have this helper and to your point it's 00:15:47.149 --> 00:15:49.370 like hey this is inefficient come along with 00:15:49.370 --> 00:15:51.190 me for this journey help me become more efficient 00:15:51.190 --> 00:15:53.129 so i think that's an interesting point you just 00:15:53.129 --> 00:15:55.570 made because we see that in real life like hey 00:15:55.570 --> 00:15:57.669 i'm trying to figure out how to do this or i 00:15:57.669 --> 00:15:59.870 you know i've never bought a business that's 00:15:59.870 --> 00:16:03.029 a business services business? What are six questions 00:16:03.029 --> 00:16:05.250 I should ask a business service business as I 00:16:05.250 --> 00:16:07.110 go to look through it? I think those are a lot 00:16:07.110 --> 00:16:08.750 of those tools. You're right. If you map those 00:16:08.750 --> 00:16:10.409 out and have a real honest conversation with 00:16:10.409 --> 00:16:12.590 yourself, they probably bubble to the top to 00:16:12.590 --> 00:16:13.710 know that that's something you need to focus 00:16:13.710 --> 00:16:15.590 on. Yeah. And I'll give you another real life 00:16:15.590 --> 00:16:17.769 example. And I think this is just the really 00:16:17.769 --> 00:16:24.049 powerful thing about, I primarily use ChatGPT. 00:16:24.639 --> 00:16:26.879 The really powerful thing about that tool is 00:16:26.879 --> 00:16:30.240 it has no limits, really. I understand that if 00:16:30.240 --> 00:16:32.139 you start plugging in some political stuff, it 00:16:32.139 --> 00:16:34.720 might not be your best friend. But I'll give 00:16:34.720 --> 00:16:38.019 you an example. Real life case study, this isn't 00:16:38.019 --> 00:16:39.580 going to go transform a mid -market business, 00:16:39.659 --> 00:16:41.799 but it certainly transformed my week a few weeks 00:16:41.799 --> 00:16:44.340 ago. I was headed to a conference, kind of a 00:16:44.340 --> 00:16:45.820 big conference, a couple thousand people were 00:16:45.820 --> 00:16:48.950 going to be there. And they encourage... people 00:16:48.950 --> 00:16:50.230 to get to know each other at this conference, 00:16:50.350 --> 00:16:52.590 lots of one on one meetings. And I got the attendee 00:16:52.590 --> 00:16:54.950 list, which was, I don't know, 1500 2000 people 00:16:54.950 --> 00:16:57.470 long, with some high level information about 00:16:57.470 --> 00:17:00.190 each each person, right name, firm, geography, 00:17:00.389 --> 00:17:02.590 like location, you know, some of their business 00:17:02.590 --> 00:17:04.869 interests, so to speak, just to keep this kind 00:17:04.869 --> 00:17:07.150 of generic as an example. I dumped that whole 00:17:07.150 --> 00:17:09.089 list into chat GPT. And I said, Hey, based on 00:17:09.089 --> 00:17:13.470 what you know about me, who are the 50 best people 00:17:13.470 --> 00:17:15.670 for me to meet at this conference. And it spit 00:17:15.670 --> 00:17:18.769 back out in, I don't know, 13 and a half seconds, 00:17:18.829 --> 00:17:21.490 maybe, you know, on the top end, it spat out 00:17:21.490 --> 00:17:23.849 the 50 people that it felt like I should get 00:17:23.849 --> 00:17:27.230 to know reasons and the logic around why I should 00:17:27.230 --> 00:17:29.250 get to know them. Then I prompted it to go chase 00:17:29.250 --> 00:17:31.029 down their email addresses because the list didn't 00:17:31.029 --> 00:17:33.450 have that. It came back with about 25 % of those 00:17:33.450 --> 00:17:35.549 that were correct. You know, so I found myself 00:17:35.549 --> 00:17:38.880 in a matter of one minute. If I were to go do 00:17:38.880 --> 00:17:40.519 that research on my own, it would have taken 00:17:40.519 --> 00:17:42.740 me weeks and weeks if I would have done it. And 00:17:42.740 --> 00:17:45.960 instead, I found myself five minutes later looking 00:17:45.960 --> 00:17:48.880 at this much more relevant list of 50, knowing 00:17:48.880 --> 00:17:51.119 exactly who I could reach out to in advance of 00:17:51.119 --> 00:17:52.980 going there because I had some email addresses. 00:17:53.180 --> 00:17:55.380 And it turned out to be very productive. Most 00:17:55.380 --> 00:17:57.059 of those folks actually that I met with were 00:17:57.059 --> 00:17:59.700 very good people for me to connect with. And 00:17:59.700 --> 00:18:02.000 I think you really just got to kind of challenge 00:18:02.000 --> 00:18:06.119 yourself to think about. your everyday or your 00:18:06.119 --> 00:18:09.539 department's daily workflows or your broader 00:18:09.539 --> 00:18:12.039 businesses pain points and where do we have a 00:18:12.039 --> 00:18:15.339 lot of humans doing things how can we do that 00:18:15.339 --> 00:18:17.819 more efficiently using this tool and quite frankly 00:18:17.819 --> 00:18:21.359 a great place to start is asking the tool for 00:18:21.359 --> 00:18:25.329 advice on how to use the tool Inception. Right? 00:18:25.390 --> 00:18:27.210 Like, hey, you know, I didn't do this with the 00:18:27.210 --> 00:18:29.410 conference example, but hey, I'm going to a conference 00:18:29.410 --> 00:18:32.130 next week. How could I use ChatGPT to help me 00:18:32.130 --> 00:18:34.509 improve my conference outcome? And I guarantee 00:18:34.509 --> 00:18:36.309 that thing's going to spit back, well, hey, I 00:18:36.309 --> 00:18:38.289 could do research in advance. I could tell you, 00:18:38.329 --> 00:18:40.390 you know, driving directions, how to get there 00:18:40.390 --> 00:18:42.910 faster. You know, a bunch of different things. 00:18:43.210 --> 00:18:44.490 And I think, you know, the more people start 00:18:44.490 --> 00:18:46.529 using that tool, the more they're going to be 00:18:46.529 --> 00:18:48.869 able to kind of understand the use cases across 00:18:48.869 --> 00:18:51.359 their businesses. And then the more they're going 00:18:51.359 --> 00:18:53.559 to go and be able to explore out there in the 00:18:53.559 --> 00:18:56.579 universe of solutions and partners and options, 00:18:56.799 --> 00:18:58.599 they're going to be better served to go in and 00:18:58.599 --> 00:19:00.819 pursue some of those. Yeah, and you touched on 00:19:00.819 --> 00:19:02.339 a couple of things in there. Maybe I've been 00:19:02.339 --> 00:19:04.000 alluding that we're going to talk about it. Maybe 00:19:04.000 --> 00:19:05.700 just jumping around the privacy side, right? 00:19:06.000 --> 00:19:07.940 I'll stick to the example you just gave, which 00:19:07.940 --> 00:19:10.539 is an excellent example. That's publicly available 00:19:10.539 --> 00:19:13.700 data that you put back into GPT to become more 00:19:13.700 --> 00:19:15.319 efficient, right? Like you didn't take anything 00:19:15.319 --> 00:19:17.640 that was of a private nature. And so one of the 00:19:17.640 --> 00:19:19.559 things I want to talk about is like, smart approaches 00:19:19.559 --> 00:19:22.339 and safe approaches to using AI. And so the example 00:19:22.339 --> 00:19:24.599 you gave is excellent. That's information that's 00:19:24.599 --> 00:19:26.740 out there. We see it on Zoom Info. You can look 00:19:26.740 --> 00:19:28.200 it up to your point if you did it manually, be 00:19:28.200 --> 00:19:30.460 going to Zoom Info and trying to find it. That's 00:19:30.460 --> 00:19:32.599 one. What we think is interesting, and if I may 00:19:32.599 --> 00:19:35.220 speak to my PE friends for a little bit, is folks 00:19:35.220 --> 00:19:39.819 are taking SIMs and loading them into GPT before 00:19:39.819 --> 00:19:42.500 a deal that's not done. That is one of those 00:19:42.500 --> 00:19:44.740 things that really concerns us that we see with 00:19:44.740 --> 00:19:46.930 our clients. To speak about Trivista, because 00:19:46.930 --> 00:19:49.329 Trivista is a business as well, is we have an 00:19:49.329 --> 00:19:52.190 internal committee of about eight folks that 00:19:52.190 --> 00:19:55.210 their committee's responsibility is to be the 00:19:55.210 --> 00:19:58.480 governing body for AI at Trivista. So it's from 00:19:58.480 --> 00:20:00.200 everything from when we use it, do we have to 00:20:00.200 --> 00:20:02.859 quote it? Which tools do we want to use? What 00:20:02.859 --> 00:20:05.160 are use cases that we're okay with? And how do 00:20:05.160 --> 00:20:07.180 we go and build future tools that make our business 00:20:07.180 --> 00:20:09.960 successful? That governing body released a position 00:20:09.960 --> 00:20:12.480 that we keep up to date on tools that are approved 00:20:12.480 --> 00:20:14.619 to be used and tools that are not. It's very 00:20:14.619 --> 00:20:16.940 intentional and it's a conversation we have and 00:20:16.940 --> 00:20:19.279 it's a continuous education. Another thing about 00:20:19.279 --> 00:20:21.019 Travista is every two weeks we have an all hands 00:20:21.019 --> 00:20:23.740 meeting and we kind of share information. AI 00:20:23.740 --> 00:20:27.599 now is on the bi -weekly share agenda because 00:20:27.599 --> 00:20:29.240 it's such an important thing to make sure we 00:20:29.240 --> 00:20:31.440 stay tight out and protected on because we deal 00:20:31.440 --> 00:20:33.319 with so much sensitive data and information. 00:20:34.440 --> 00:20:36.619 Outside of that, another way to use it pretty 00:20:36.619 --> 00:20:39.819 intelligently is the same way you hear this phrase 00:20:39.819 --> 00:20:41.559 all the time, crawl, walk, run, right? Start 00:20:41.559 --> 00:20:44.380 slow and then build up. To Mike's point around, 00:20:44.619 --> 00:20:47.950 hey. Where do I find things that I can do, do 00:20:47.950 --> 00:20:49.890 a little test and get better and better? You 00:20:49.890 --> 00:20:51.750 may at one point, now that you go to enough conferences, 00:20:51.849 --> 00:20:54.450 you may go and build a custom GPT yourself that 00:20:54.450 --> 00:20:56.970 does that conference work. But asking questions, 00:20:57.230 --> 00:20:59.269 exploring, working through it is a better way 00:20:59.269 --> 00:21:01.410 to start. One of the things I always encourage 00:21:01.410 --> 00:21:05.589 people is most of the tools today use OpenAI 00:21:05.589 --> 00:21:08.240 on the back end. So if you go into ChatGPT and 00:21:08.240 --> 00:21:10.079 do a little experiment and it's off limits, it 00:21:10.079 --> 00:21:12.119 doesn't help, it doesn't work. The likelihood 00:21:12.119 --> 00:21:15.099 of you creating a commercial tool at scale that 00:21:15.099 --> 00:21:17.319 would be able to solve that same issue is low. 00:21:17.940 --> 00:21:20.319 One of the things that we've done at Trivista 00:21:20.319 --> 00:21:22.140 to protect ourselves is we went out and got something 00:21:22.140 --> 00:21:26.740 called Azure AI. Azure AI runs on OpenAI, which 00:21:26.740 --> 00:21:30.640 is the group that owns ChatGPT. That allowed 00:21:30.640 --> 00:21:34.000 us to create a very safe, let's call it enclave, 00:21:34.119 --> 00:21:36.200 where we're able to load data in a secure way. 00:21:36.319 --> 00:21:38.700 That data doesn't go out to the web and it doesn't 00:21:38.700 --> 00:21:41.420 train models. It's only consuming the data that's 00:21:41.420 --> 00:21:43.359 out there. And so you hear this thing of like 00:21:43.359 --> 00:21:46.240 large language models or LLMs. It uses the large 00:21:46.240 --> 00:21:48.500 language model and then looks at our information 00:21:48.500 --> 00:21:51.339 and helps us align to the two of them. One of 00:21:51.339 --> 00:21:53.119 the things that we've been very careful about 00:21:53.119 --> 00:21:55.839 is ensuring that that data is protected. If you're 00:21:55.839 --> 00:21:57.900 a mid -market organization that wants to go build 00:21:57.900 --> 00:22:00.259 a really neat tool and you feel that you've done 00:22:00.259 --> 00:22:02.220 the small test, you've done the things that we've 00:22:02.220 --> 00:22:04.039 said, you think there's something important to 00:22:04.039 --> 00:22:07.759 do, then that is the time that is to go look 00:22:07.759 --> 00:22:09.960 at these tools that are more enterprise and robust. 00:22:10.880 --> 00:22:12.980 To Mike's earlier point, there's a heck of a 00:22:12.980 --> 00:22:16.259 cost when you go down that way. So the last point 00:22:16.259 --> 00:22:19.160 we'll make about this is focus on... business 00:22:19.160 --> 00:22:21.019 first and making sure there's a business case 00:22:21.019 --> 00:22:23.319 that drives value. You hear Mike and I say all 00:22:23.319 --> 00:22:26.039 the time on the way to grow EBITDA is have a 00:22:26.039 --> 00:22:28.240 business case that creates value that you can 00:22:28.240 --> 00:22:30.900 stand behind with KPIs. The same needs to be 00:22:30.900 --> 00:22:33.420 applied when you think about AI. I'm going to 00:22:33.420 --> 00:22:36.619 go solve X issue. It costs me Y. The use case 00:22:36.619 --> 00:22:39.549 is Z. Those type of conversations need to happen 00:22:39.549 --> 00:22:42.069 around AI. It can be fun. Boards may pressure 00:22:42.069 --> 00:22:44.069 you. Friends may tell you they're doing things. 00:22:44.289 --> 00:22:46.730 But we'll say you still need to follow the basic 00:22:46.730 --> 00:22:48.670 principles just like you do with anything else. 00:22:50.089 --> 00:22:52.150 Anything else you want to add before we wrap 00:22:52.150 --> 00:22:54.369 up? No, I think that was great. I mean, obviously, 00:22:54.410 --> 00:22:59.630 the privacy thing was a huge concern 12, 18 months 00:22:59.630 --> 00:23:03.750 ago. And I think that what we've seen is ChatGPT 00:23:03.750 --> 00:23:06.970 and Copilot, et cetera, come out with. versions 00:23:06.970 --> 00:23:09.650 of itself that are actually protecting data and 00:23:09.650 --> 00:23:11.089 things like that. And I think everybody was a 00:23:11.089 --> 00:23:13.349 little bit skeptical at first. I think everybody's 00:23:13.349 --> 00:23:15.250 kind of backing off that a little bit. But you 00:23:15.250 --> 00:23:16.710 do got to make sure that you're using an environment 00:23:16.710 --> 00:23:18.589 and you're loading data in that you feel good 00:23:18.589 --> 00:23:21.329 about. And obviously, don't go use the public 00:23:21.329 --> 00:23:24.549 version of these tools and dump all your company's 00:23:24.549 --> 00:23:26.809 most sensitive data into them without thinking 00:23:26.809 --> 00:23:29.750 about it. So yeah, great feedback. So James, 00:23:29.869 --> 00:23:35.160 before we close here, I got one for you. What 00:23:35.160 --> 00:23:38.420 year was the term artificial intelligence first 00:23:38.420 --> 00:23:44.660 coined? Okay, so I feel like this is one that 00:23:44.660 --> 00:23:46.759 I'm going to get very wrong based on the tee 00:23:46.759 --> 00:23:48.500 up because you chose this. And I know you're 00:23:48.500 --> 00:23:50.799 coming off a win where you got one right. So 00:23:50.799 --> 00:23:54.319 it's one correct and 11 wrong. So I'm sure this 00:23:54.319 --> 00:23:56.680 is going to be tough. I'm going to go, I feel 00:23:56.680 --> 00:23:58.599 like the 80s was a big explosion for technology. 00:23:58.599 --> 00:24:01.400 Are you an 80s baby? Apple 1984 commercial. You 00:24:01.400 --> 00:24:02.779 are. See, that's why you thought that it was 00:24:02.779 --> 00:24:06.049 the 80s. You know, James came on online. I am. 00:24:06.130 --> 00:24:09.690 Now he's a technologist. You couldn't, it'd be 00:24:09.690 --> 00:24:16.150 tough to be further away. 1956, at the Dartmouth 00:24:16.150 --> 00:24:18.690 Conference, which I do not know what the Dartmouth 00:24:18.690 --> 00:24:21.430 Conference is, but apparently, allegedly, 1956 00:24:21.430 --> 00:24:24.470 at the Dartmouth Conference just goes to show 00:24:24.470 --> 00:24:28.049 you that AI isn't quite as new as everybody thinks. 00:24:28.329 --> 00:24:31.740 To me, it will always be. ChatGPT's public launch 00:24:31.740 --> 00:24:34.180 date, which was like November 23, if I'm not 00:24:34.180 --> 00:24:37.180 mistaken, changed my life for the better, me 00:24:37.180 --> 00:24:40.200 and everybody else on earth. But yeah, 1956. 00:24:41.759 --> 00:24:44.019 Well, I can't believe it's taken us this long 00:24:44.019 --> 00:24:46.200 to figure out what really adds value and what 00:24:46.200 --> 00:24:48.240 really is, I guess. It's a great term to come 00:24:48.240 --> 00:24:50.819 up, but I, like you, I can't imagine my life 00:24:50.819 --> 00:24:52.759 without it now, from distilling data to writing 00:24:52.759 --> 00:24:55.319 emails to researching things to even just trying 00:24:55.319 --> 00:24:57.619 to figure out. What do I do in certain situations 00:24:57.619 --> 00:25:00.059 with having a baby? It's been a great tool and 00:25:00.059 --> 00:25:02.460 a huge help on personal and a professional level. 00:25:03.480 --> 00:25:05.819 Definitely, definitely. Well, thanks everybody 00:25:05.819 --> 00:25:08.279 for tuning in and we'll catch you next time. 00:25:08.339 --> 00:25:12.140 Don't forget to follow us, like us, subscribe 00:25:12.140 --> 00:25:15.440 to us. If you're gonna DM us, DM James. He's 00:25:15.440 --> 00:25:17.700 kind of more hip to that, he's more hip to the 00:25:17.700 --> 00:25:21.559 DMs, but we'll see you next time. Thanks for 00:25:21.559 --> 00:25:24.359 tuning into this episode of Growing EBITDA. If 00:25:24.359 --> 00:25:27.190 you liked this episode, hit subscribe, or follow 00:25:27.190 --> 00:25:30.029 us on LinkedIn for updates. Got a topic you'd 00:25:30.029 --> 00:25:32.750 like us to cover? Drop us a message. We'd love 00:25:32.750 --> 00:25:33.309 to hear from you.