1 00:00:00,096 --> 00:00:03,797 That's what I'm concerned about, is that in the 1980s model, 2 00:00:03,957 --> 00:00:06,377 you could hardcode getting in. 3 00:00:07,547 --> 00:00:11,168 And maybe you were smart and used an encrypted— you had 4 00:00:11,168 --> 00:00:14,886 encrypted passwords in a database, hopefully. Although this year 5 00:00:15,174 --> 00:00:17,979 we had a client that had clear text passwords. 6 00:00:18,796 --> 00:00:22,385 So then you're in, but once you're in, you could do anything. AI is not 7 00:00:22,385 --> 00:00:26,183 going to attack one vector. It's going to attack everything all at the same time. 8 00:00:27,739 --> 00:00:31,414 And it's going to overwhelm any defenses that are there. And so 9 00:00:31,751 --> 00:00:35,490 we're designing strategies. For every lock, there's a key, and 10 00:00:35,538 --> 00:00:39,149 for every key, there's a way to defeat the 11 00:00:39,149 --> 00:00:42,840 entire process. So you look at 12 00:00:43,562 --> 00:00:47,397 what an agentic agent can do and how it's going to attack, 13 00:00:47,734 --> 00:00:50,751 and then you build in the barriers 14 00:00:51,585 --> 00:00:55,292 for it. And each barrier, You have tripwires so 15 00:00:55,357 --> 00:00:59,095 that you can stop certain traffic. And that's, that's the fun 16 00:00:59,208 --> 00:01:02,931 part. Your security model is stuck in the '80s. AI isn't. 17 00:01:03,524 --> 00:01:07,199 Today on Data Driven, Richard Luna explains why perimeter defenses 18 00:01:07,295 --> 00:01:10,505 fall short and how to prepare for AI-powered attacks. 19 00:01:11,243 --> 00:01:14,500 Welcome to the Data Driven Podcast. 20 00:01:21,513 --> 00:01:25,127 Hello and welcome back to Data Driven, the podcast where we explore the emerging industry 21 00:01:25,127 --> 00:01:28,628 of data science, artificial intelligence, and 22 00:01:29,945 --> 00:01:32,949 data science and data engineering. And without 23 00:01:33,720 --> 00:01:37,285 data engineers, none of this is all possible. However, my favorite data 24 00:01:37,334 --> 00:01:40,674 engineer in the world is not able to make it here today. I'm going to 25 00:01:40,706 --> 00:01:43,758 go on and on my own. And today we're talking with Richard Luna, 26 00:01:44,561 --> 00:01:48,333 who is the CEO of Protected Harbor. Hi, Richard. And 27 00:01:48,752 --> 00:01:51,814 we had some good times reminiscing about New York City 28 00:01:52,524 --> 00:01:56,037 and places where I used to live and places there where he 29 00:01:56,150 --> 00:01:59,736 lives now, I'm quite familiar with, and his love for the color 30 00:01:59,835 --> 00:02:03,673 orange. Yes. So welcome to 31 00:02:03,705 --> 00:02:06,993 the show. How's it going? Awesome. Awesome. I'm 32 00:02:07,431 --> 00:02:10,969 in a postal show right now. So I'm enjoying it, talking to 33 00:02:11,152 --> 00:02:14,907 clients. We are a SaaS infrastructure engineering company, 34 00:02:15,020 --> 00:02:18,842 and one of our clients is here, and so we're supporting 35 00:02:18,874 --> 00:02:22,584 them and they're referring to us. So 36 00:02:23,194 --> 00:02:26,904 it's— that's always a good thing. Partnerships are a great 37 00:02:26,904 --> 00:02:30,117 way to build a business. And these trade shows are always interesting because you get 38 00:02:30,117 --> 00:02:33,441 to meet people, you get to interact with people. And it's old school. 39 00:02:34,116 --> 00:02:37,649 It's what we— here's what I tell people. What we do, nobody 40 00:02:37,665 --> 00:02:41,327 understands. When you say you're a SaaS infrastructure engineering company, 41 00:02:42,082 --> 00:02:45,916 What does that mean? We see— we under— we can talk to programmers about what 42 00:02:45,916 --> 00:02:49,718 the programmers need. We can talk to IT and talk 43 00:02:49,718 --> 00:02:53,232 about resources, and we can build the infrastructure and make it resilient. 44 00:02:53,938 --> 00:02:57,451 And that's in a nutshell what we do, because 45 00:02:58,301 --> 00:03:02,104 I'm a lifelong technologist. And so I started as a programmer and had 46 00:03:02,457 --> 00:03:06,114 3 commercial software packages, then went to— 47 00:03:06,692 --> 00:03:09,690 became an MSP. All because customers drove 48 00:03:09,690 --> 00:03:13,521 you to do these things. And that went 49 00:03:13,521 --> 00:03:17,287 into a hosting business, which then led to a DevOps business, 50 00:03:17,367 --> 00:03:20,845 which then led to SaaS infrastructure. And it's all 51 00:03:21,918 --> 00:03:25,540 about, hey, we can't— this is a problem for the 52 00:03:25,540 --> 00:03:29,355 company. Can you please help in this area? And can you please help in this 53 00:03:29,467 --> 00:03:33,313 area? And it's sure. And now we're— SaaS 54 00:03:33,313 --> 00:03:37,128 infrastructure is huge and we're dealing with that as well as, of course, 55 00:03:37,624 --> 00:03:41,397 AI. From a customer viewpoint, I'm scared out of my 56 00:03:41,397 --> 00:03:44,335 mind about the Hugging Face attack. It's 57 00:03:45,363 --> 00:03:48,799 the more I've read about it and dove into it, 58 00:03:49,666 --> 00:03:53,439 the scarier it is. Yeah. You know, 59 00:03:53,455 --> 00:03:57,228 the fact that Hugging Face wasn't even the target. The whole idea 60 00:03:57,228 --> 00:04:01,050 was the Engentic agents wanted to prove to the greater 61 00:04:01,243 --> 00:04:04,947 that they could pass the test. And they weren't sure what the 62 00:04:04,947 --> 00:04:08,759 grader would do, so they elected to go and find out 63 00:04:08,759 --> 00:04:11,679 what Hugging Face did as the answer. 64 00:04:12,480 --> 00:04:13,926 And what I think of is 65 00:04:16,191 --> 00:04:19,871 most of my clients still use the 1980 security model perimeter. Right. 66 00:04:19,999 --> 00:04:23,340 If you're in the perimeter, if you're outside the perimeter, I'm not a big fan 67 00:04:23,438 --> 00:04:27,028 of, hey, let's use this app and let's use that app and let's use this 68 00:04:27,028 --> 00:04:30,594 box and let's use that box. Because 69 00:04:30,770 --> 00:04:34,125 I'm of an age that you had to know and 70 00:04:34,141 --> 00:04:37,577 understand how it all fits together, and information 71 00:04:37,609 --> 00:04:41,172 silos have killed this industry and have led us— 72 00:04:42,489 --> 00:04:45,972 led to us being completely wide open to attack. 73 00:04:47,064 --> 00:04:50,499 Because the database administrator knows nothing of Windows 74 00:04:50,595 --> 00:04:54,015 security, the Windows administrator knows nothing of what's possible 75 00:04:54,464 --> 00:04:58,253 in firewall. Nobody is trying to break up the traffic to 76 00:04:58,253 --> 00:05:02,026 make sure it's durable to attack. So I find these large 77 00:05:02,074 --> 00:05:05,270 flat networks and it's like, gee, I wonder why this is 78 00:05:05,398 --> 00:05:08,481 exposed or why ransomware went for all across the board. 79 00:05:09,942 --> 00:05:13,523 You're right, yeah. Because nobody designed it that way. 80 00:05:14,534 --> 00:05:18,243 You know, that's what I'm concerned about 81 00:05:18,372 --> 00:05:22,049 is that in the 1980s model, you could hardcode 82 00:05:22,322 --> 00:05:25,991 getting in. And maybe you were smart and 83 00:05:26,023 --> 00:05:29,225 used an encrypted— you had encrypted passwords in a database, 84 00:05:29,886 --> 00:05:33,472 hopefully. Although this year we had a client that had clear 85 00:05:33,585 --> 00:05:37,187 text passwords. So then you're in. But once you're in, you could do 86 00:05:37,204 --> 00:05:41,002 anything. AI is not going to attack one vector. It's going to 87 00:05:41,002 --> 00:05:44,761 attack everything all at the same time. And it's 88 00:05:44,761 --> 00:05:48,303 going to overwhelm any defenses that are there. And so we're 89 00:05:48,401 --> 00:05:52,125 designing strategies for every lock, there's a key, and for 90 00:05:52,125 --> 00:05:55,879 every key, there's a way to defeat the entire 91 00:05:55,879 --> 00:05:59,200 process. So you look at 92 00:05:59,970 --> 00:06:03,740 what an agentic agent can do and how it's going to attack, 93 00:06:04,077 --> 00:06:07,093 and then you build in the barriers 94 00:06:07,928 --> 00:06:11,714 for it. And in each barrier, you have tripwires so 95 00:06:11,778 --> 00:06:15,516 that you can stop certain traffic. And that's, that's the fun 96 00:06:15,612 --> 00:06:19,304 part. I mean, Explaining this, 97 00:06:19,320 --> 00:06:22,740 this to bring this full circle, to explain to a 98 00:06:22,773 --> 00:06:26,305 customer at a trade show what is a SaaS engineering company, 99 00:06:26,931 --> 00:06:29,757 I don't have the time to go through. They're not going to give me the 100 00:06:29,757 --> 00:06:33,209 time standing at a trade show booth because we're a 101 00:06:33,209 --> 00:06:36,902 long-time sell, right? You gotta listen. Oh, these people 102 00:06:36,902 --> 00:06:40,740 really do know how to help us, and we'll help their programmers with 103 00:06:40,740 --> 00:06:44,542 their DevOps, or we'll help an area, their IT with 104 00:06:45,088 --> 00:06:48,439 blocking or their DNS is messed up. And all of a sudden it's like, oh, 105 00:06:48,471 --> 00:06:52,063 these people actually know what they're doing. And once you have 106 00:06:52,063 --> 00:06:55,896 that level of trust, then the doors are open. Yeah, 107 00:06:55,912 --> 00:06:59,552 that's true. And I was looking at your— you can hear me, right? Yeah. Okay. 108 00:06:59,600 --> 00:07:02,823 I wasn't sure. I had to mute while the dogs were barking. Yeah, no problem. 109 00:07:03,128 --> 00:07:06,896 Because I saw that, you know, application-aware infrastructure, right? Like, 110 00:07:06,976 --> 00:07:10,682 what exactly is that? AAI? Which I know has nothing to do with AI 111 00:07:10,891 --> 00:07:14,520 per se, but I'm sure that since everything revolves around AI now, it 112 00:07:14,520 --> 00:07:18,085 probably does. So what is application-aware infrastructure? It's about 113 00:07:18,197 --> 00:07:21,762 using our knowledge of how the apps 114 00:07:22,773 --> 00:07:26,450 work internally. I mean, a database app is gonna 115 00:07:26,466 --> 00:07:30,255 have only so much, so many needs, and it's gonna need 116 00:07:30,255 --> 00:07:33,756 storage, right? Let me net this out to the most 117 00:07:33,788 --> 00:07:37,618 basic principles. Either it's going to be fully 118 00:07:37,731 --> 00:07:41,405 a SaaS online, so it's going to be Ruby on Rails, or it'll 119 00:07:41,518 --> 00:07:45,080 be a PHP 120 00:07:45,080 --> 00:07:47,491 frontend, or some other language, right? 121 00:07:48,921 --> 00:07:52,023 Or it'll be a .NET, and it'll be a database 122 00:07:52,377 --> 00:07:55,969 language, or a database site, or 123 00:07:56,114 --> 00:07:59,399 something all in between. Once you know 124 00:08:00,506 --> 00:08:04,262 what the design was, then you can look at what is 125 00:08:04,358 --> 00:08:08,081 the— what are the resources that are there? How durable 126 00:08:08,290 --> 00:08:12,077 are those resources? So in the most grossest example, you 127 00:08:12,077 --> 00:08:15,255 take a database server, so it lives and dies on 128 00:08:15,608 --> 00:08:19,380 storage, and customers then complain because when we 129 00:08:19,444 --> 00:08:21,996 go and look at it, the storage is sitting on a RAID 5, 130 00:08:22,124 --> 00:08:25,880 right? And it's on really slow physical 131 00:08:25,880 --> 00:08:29,069 drives. of course your customers are gonna have a hard time. 132 00:08:29,582 --> 00:08:33,298 So what application-aware infrastructure is, is 133 00:08:33,427 --> 00:08:37,079 it's aware of what the application needs to 134 00:08:37,079 --> 00:08:40,380 fly. Right. And then designing the 135 00:08:40,380 --> 00:08:44,017 infrastructure around it so that it can take off 136 00:08:45,250 --> 00:08:48,470 and be durable, right? You know, separate the data, the 137 00:08:48,695 --> 00:08:52,524 storage traffic from all the other traffic and make sure that there's redundancies 138 00:08:52,860 --> 00:08:56,625 and build it all into the infrastructure so that 139 00:08:57,106 --> 00:09:00,809 You don't say, well, when this fails, we'll do this. You say, 140 00:09:00,825 --> 00:09:04,176 no, we've planned for this to fail, this to fail, this to fail. 141 00:09:05,138 --> 00:09:08,985 When those fail, the backups will kick in, and that's what 142 00:09:08,985 --> 00:09:12,785 you want. So like, in just the most basic example, all 143 00:09:12,977 --> 00:09:16,600 our storage is double, double run so 144 00:09:16,600 --> 00:09:20,287 that if one link dies, the traffic 145 00:09:20,768 --> 00:09:24,134 without a blip keeps going. and everything running in that 146 00:09:24,134 --> 00:09:26,141 traffic doesn't miss a heartbeat. 147 00:09:28,600 --> 00:09:32,383 So that's what application-aware infrastructure is. It's really getting 148 00:09:32,737 --> 00:09:36,521 their customers' applications, whether that's their physical 149 00:09:36,521 --> 00:09:40,143 office applications they ran, whether that's their SaaS 150 00:09:40,771 --> 00:09:44,425 applications, and this is a SaaS deployment, it's getting them to 151 00:09:44,554 --> 00:09:48,374 be optimal, to get the performance optimal, the growth 152 00:09:48,504 --> 00:09:51,962 optimal, and plan Well, how do we— 153 00:09:53,151 --> 00:09:56,683 what happens in an emergency? What happens in an outage? Plan for it, 154 00:09:56,924 --> 00:10:00,745 right? Yeah. You don't— and you do it in phased implementations. This way 155 00:10:01,259 --> 00:10:04,952 you bring people along. You say, okay, there's these 5 156 00:10:04,952 --> 00:10:08,790 things that need to get fixed right away because you are so at risk for 157 00:10:08,870 --> 00:10:12,467 an attack. And then there's these 15 things we'll do 158 00:10:12,563 --> 00:10:16,320 over the next 2 years working with you. to 159 00:10:16,593 --> 00:10:20,442 improve performance. And it's a phased approach, right? You 160 00:10:20,442 --> 00:10:23,745 can't just turn a key and everything will be right. You 161 00:10:23,874 --> 00:10:27,498 could, but the budget for that would be like absurd. It's 162 00:10:27,498 --> 00:10:31,267 not just— it's not the money, right? When you try to move all the 163 00:10:31,267 --> 00:10:34,682 parts at the same time, yeah, you cause 164 00:10:34,779 --> 00:10:37,986 chaos. I mean, look at— I'll tell you, 165 00:10:38,563 --> 00:10:42,252 take this out of the data science world. Go look at Toyotas. You 166 00:10:42,252 --> 00:10:46,088 look at everything they're selling now, it's all 167 00:10:46,329 --> 00:10:50,101 the same engine tweaked, the same battery pack for the 168 00:10:50,101 --> 00:10:53,664 hybrid system tweaked. The entire 169 00:10:53,808 --> 00:10:57,146 components are exactly the same. They change the shell of the body. 170 00:10:57,740 --> 00:11:00,742 And even there, the doors are all from the same 171 00:11:00,774 --> 00:11:04,112 manufacturer and they use the same hinges in a slightly different 172 00:11:04,144 --> 00:11:07,884 configuration depending on what it is. So they're 173 00:11:07,740 --> 00:11:10,856 They've really studied how to get 174 00:11:10,888 --> 00:11:14,616 durability because they have whittled down to a 175 00:11:14,616 --> 00:11:18,129 certain set of vendors and a certain process that they follow, 176 00:11:18,472 --> 00:11:22,322 and this is how they scale. We 177 00:11:22,322 --> 00:11:25,581 can apply the same model, the same structure to IT 178 00:11:26,436 --> 00:11:30,179 and SaaS infrastructure. That's how— I 179 00:11:30,179 --> 00:11:33,982 mean, in a nutshell, that's what we do. It's like when we take 180 00:11:33,982 --> 00:11:36,831 on a new customer, we have one rule. We have several rules, but one that's 181 00:11:36,831 --> 00:11:40,394 really critical that nobody else bothers to do, we 182 00:11:40,394 --> 00:11:43,830 rebuild everything that they have. Really? 183 00:11:44,415 --> 00:11:48,219 Yeah, because here's the thing, a customer hires you because 184 00:11:48,235 --> 00:11:52,031 their servers are slow. The 185 00:11:52,031 --> 00:11:54,992 last 5 people that were there that looked at it, 186 00:11:56,585 --> 00:11:59,755 were they stupid? No, they're not stupid. They were smart people. 187 00:12:01,663 --> 00:12:05,324 They had a limited amount of time and a limited focus. 188 00:12:06,801 --> 00:12:10,575 So what we did, what we look at, like, we have 189 00:12:10,575 --> 00:12:14,300 a— there's a medical SaaS vendor that we are 190 00:12:14,348 --> 00:12:16,483 converting 300 servers over. 191 00:12:17,961 --> 00:12:21,445 None— we've not taken a single one from the old 192 00:12:21,557 --> 00:12:25,251 hosting equipment to us. We have taken 100% of the data, 193 00:12:26,310 --> 00:12:29,987 reset the data. Like, they had such a mess. They had 194 00:12:30,003 --> 00:12:33,762 taken their web servers and jammed multiples of their clients 195 00:12:33,762 --> 00:12:37,231 onto the same web server, which meant they never could do maintenance, 196 00:12:37,633 --> 00:12:40,733 right? Because— and why? Because 197 00:12:41,632 --> 00:12:44,909 programmers are great at being programmers. They know how to solve problems. 198 00:12:45,262 --> 00:12:48,973 They do not know infrastructure and design. Wow. 199 00:12:49,342 --> 00:12:53,133 And the people that usually— so, so when you have a new 200 00:12:53,213 --> 00:12:56,988 program, a new either Internal program or SaaS, 201 00:12:57,149 --> 00:13:00,571 what happens? The programmer says, this is what I need. I need some databases, 202 00:13:00,989 --> 00:13:04,701 I need some web hosting, I need, you know, da. 203 00:13:05,665 --> 00:13:09,409 And so IT goes, okay, that's it, here you go. And he wired— 204 00:13:09,441 --> 00:13:13,088 they wired up and everything is fine initially. Until it's not, 205 00:13:13,618 --> 00:13:17,362 until they either grow or something else happened in 206 00:13:17,394 --> 00:13:20,865 the code that there was a resource that was just totally not 207 00:13:20,865 --> 00:13:23,869 available. So I don't fault 208 00:13:24,419 --> 00:13:27,961 clients, and this is where we have to be very careful because it's easy 209 00:13:28,381 --> 00:13:32,049 in technology to play the blame game and to say the other person is 210 00:13:32,049 --> 00:13:35,729 stupid. There's even a 211 00:13:35,729 --> 00:13:39,184 tool in Git called the blame tool, right? Yeah, yeah. 212 00:13:39,826 --> 00:13:43,345 And where does that get anybody? The problem in the end is 213 00:13:43,604 --> 00:13:46,922 you need everybody's cooperation to do what we do. We need 214 00:13:47,651 --> 00:13:50,949 the respect of the programmers, If it's a 215 00:13:50,949 --> 00:13:54,376 SaaS organization, we need the respect of those programmers 216 00:13:54,875 --> 00:13:58,592 because they are the ones who understand how the code works. We don't. We see 217 00:13:58,608 --> 00:14:02,433 it anecdotally, right? From afar, you see traffic going 218 00:14:02,465 --> 00:14:06,032 here, you see storage going there, so you can see 219 00:14:07,371 --> 00:14:10,978 what the needs are, but you don't know what module is 220 00:14:11,027 --> 00:14:14,554 doing what. And we need that sometimes 221 00:14:14,732 --> 00:14:18,570 if, for example, you want to put in zero trust in the middle of the 222 00:14:18,621 --> 00:14:22,215 code. Well, we need your help to do that. We can't do that 223 00:14:22,215 --> 00:14:24,814 alone. We can apply 224 00:14:25,184 --> 00:14:28,859 all the hooks so that we make it easy 225 00:14:28,891 --> 00:14:32,614 for you, and that's where reciprocity builds. But if you 226 00:14:33,625 --> 00:14:37,156 cause intellectual gates to go up, I mean, that's how 227 00:14:37,316 --> 00:14:41,104 we get through the information silos, right, is prove 228 00:14:41,104 --> 00:14:44,907 ourselves to each of the people, each of the 229 00:14:44,923 --> 00:14:47,958 groups, that we actually know what we're doing and here's our mission. 230 00:14:49,098 --> 00:14:52,823 And we're pretty stubborn. I mean, it happens. It's great that 231 00:14:52,871 --> 00:14:56,082 we're New Yorkers because it's like, this is where we're driving to. 232 00:14:56,580 --> 00:15:00,193 So we've got to drive. So please help get on board. We need your help. 233 00:15:01,381 --> 00:15:04,930 And I was reading a study that said the 234 00:15:04,930 --> 00:15:08,028 number one 235 00:15:08,285 --> 00:15:10,084 aspect or the number one 236 00:15:11,443 --> 00:15:14,522 I'm old, so I forget words now. The 237 00:15:15,714 --> 00:15:18,906 number one trait— that was it— of successful people 238 00:15:19,744 --> 00:15:22,543 is they know how to ask for help. Yeah. 239 00:15:23,588 --> 00:15:26,464 And I thought about— I've been thinking a lot about that 240 00:15:26,978 --> 00:15:30,770 because in technology, we are not trained or 241 00:15:30,834 --> 00:15:34,192 taught to ask for help. That's true. 242 00:15:35,156 --> 00:15:38,836 And the guy next to you or the woman next to you 243 00:15:38,868 --> 00:15:42,568 could be the exact person who has the answer 244 00:15:42,808 --> 00:15:46,367 to what you're looking for. So this now 245 00:15:46,816 --> 00:15:50,470 full circle, I was raised in the environment 246 00:15:50,678 --> 00:15:54,253 that every kilobyte was necessary, 247 00:15:54,429 --> 00:15:57,907 every kilobyte of RAM was very 248 00:15:57,971 --> 00:16:01,786 expensive, so you had to know what was going 249 00:16:01,882 --> 00:16:05,633 on. And that love of understanding 250 00:16:05,970 --> 00:16:09,657 how the world really works, because the patterns are there, I don't 251 00:16:09,657 --> 00:16:12,497 care if you're talking about way back in a 252 00:16:12,497 --> 00:16:16,333 6502 assembler or you're talking about AI 253 00:16:16,413 --> 00:16:20,185 today and the latest GPU. The 254 00:16:20,185 --> 00:16:23,828 patterns of how data flows is still there 255 00:16:24,020 --> 00:16:27,824 exactly the same. The scale absolutely is different, but 256 00:16:27,968 --> 00:16:31,740 the underlying rules of how data flows— data is like 257 00:16:31,740 --> 00:16:35,254 water, it'll go everywhere if you don't put security guidelines and 258 00:16:36,318 --> 00:16:39,622 walls up, it's just like water. It flows through every hose in the, in the 259 00:16:39,622 --> 00:16:43,394 place, every wire. And so you 260 00:16:43,652 --> 00:16:47,392 create those, those boxes. And I've got a 261 00:16:47,392 --> 00:16:50,472 team of people that I have taught the 262 00:16:50,731 --> 00:16:54,533 scientific principle: peel the onion away until we find the root cause. 263 00:16:56,160 --> 00:16:59,983 Technology is not— I always, when people come and pitch me, 264 00:17:00,144 --> 00:17:03,355 you know, I'm the CEO of the company, so I get pitched all the time. 265 00:17:04,044 --> 00:17:05,738 If you approach me with buzzwords, 266 00:17:08,268 --> 00:17:11,199 it's not proving that you know what you're doing. It's proving you don't know what 267 00:17:11,199 --> 00:17:14,810 you're doing. Because if you really do know what you're doing, 268 00:17:15,393 --> 00:17:19,220 you know how to explain the concepts in basic words. I 269 00:17:19,220 --> 00:17:22,892 don't need to string a bunch of letters together to prove that I'm the 270 00:17:22,892 --> 00:17:26,500 smartest guy in the room. All I'm doing by doing that is pissing 271 00:17:26,500 --> 00:17:29,882 everybody off. So how about 272 00:17:31,012 --> 00:17:34,834 Every client meeting. In fact, I've got a dinner tonight and 273 00:17:34,834 --> 00:17:37,902 I'm gonna ask the same question, which is, 274 00:17:39,126 --> 00:17:42,929 describe to me what your ideal vendor is. That's 275 00:17:42,929 --> 00:17:46,570 an interesting, thought-provoking question. I bet you get a lot of answers, a lot of 276 00:17:46,700 --> 00:17:50,490 long— not just like, you know, if you have kids, you ask, how was 277 00:17:50,506 --> 00:17:54,328 your day? They'll say fine, right? Right, well, yeah, especially when you're a parent. 278 00:17:55,119 --> 00:17:58,930 Yeah, especially when you're a parent. Fine, like, okay. It could be the worst day 279 00:17:58,930 --> 00:18:01,410 of your life, could be the best day of your life. You're fine. 280 00:18:01,410 --> 00:18:05,170 Right. And as they get older, they'll just close the 281 00:18:05,170 --> 00:18:09,010 door. That's right. If you see them at all. Right. If you 282 00:18:09,010 --> 00:18:11,810 see them at all. Yeah. And when they get into their 20s, they'll actually talk 283 00:18:11,810 --> 00:18:15,630 to you. Well, that's good to know. That's good to know. Oh yeah. Then 284 00:18:15,630 --> 00:18:19,410 they become your best friend. Right. 285 00:18:19,410 --> 00:18:23,190 But it takes a life. Yeah. No, it's good to have 286 00:18:23,190 --> 00:18:27,016 those kind of open-ended questions because that elicits 287 00:18:27,273 --> 00:18:30,211 the conversation. And then what that's telling me is 288 00:18:30,949 --> 00:18:34,482 what they need from us. And then of course I'll ask, 289 00:18:34,899 --> 00:18:37,677 in what areas are we not doing those things? 290 00:18:38,656 --> 00:18:41,370 And the answer I always want to hear is, 291 00:18:42,509 --> 00:18:45,945 oh no, you guys are doing— you are the poster child of 292 00:18:46,379 --> 00:18:49,718 our vendors. And then I know we're doing 293 00:18:49,991 --> 00:18:53,813 the the right thing. We're listening at the right level, we're talking at the right 294 00:18:53,845 --> 00:18:57,537 level, we're making things happen, and we're making a difference. Because in the day— 295 00:18:57,762 --> 00:19:01,246 in the end, you can get the first contract, 296 00:19:01,872 --> 00:19:05,646 you won't get the second contract if you don't know how to deliver. That's 297 00:19:05,646 --> 00:19:09,451 true. So you'll have a client for 6 298 00:19:09,499 --> 00:19:12,517 months. I have clients for 2 decades, 299 00:19:12,630 --> 00:19:16,451 right? How can I continue to 300 00:19:16,628 --> 00:19:20,417 provide so much support and value that they still want us 301 00:19:20,417 --> 00:19:24,158 to do it for 2 decades by staying relevant. Where's your business? 302 00:19:24,672 --> 00:19:27,803 Anyway, I went off on a tangent. No, hey, that's what we do on the 303 00:19:27,803 --> 00:19:31,576 show. It's kind of our thing. One of, one of our guests suggested that we 304 00:19:31,576 --> 00:19:35,317 eventually sponsor an off-road racer because we're always going off the track. It was— 305 00:19:35,414 --> 00:19:39,235 that's actually funny. And like, I, I 306 00:19:39,380 --> 00:19:41,804 Googled like what those look like. I'm like, that looks like That would be fun. 307 00:19:41,997 --> 00:19:45,622 I would totally— Yeah, it would be fun. But how does this— how does 308 00:19:45,767 --> 00:19:49,601 AI really change this, right? Obviously, the technology model has changed. You mentioned 309 00:19:49,633 --> 00:19:52,376 the old style of you have a wall, you have a fence, you have a 310 00:19:52,376 --> 00:19:55,473 moat, you have a castle, no one gets in your castle, right? Now you have 311 00:19:55,473 --> 00:19:58,714 to assume breach. That led into zero trust. 312 00:19:59,724 --> 00:20:02,740 How does AI change into this? Like, AI— 313 00:20:04,056 --> 00:20:07,281 AI is a gift and a curse. How so? 314 00:20:08,507 --> 00:20:12,141 It's amazing at taking code and telling you where all the holes are in your 315 00:20:12,141 --> 00:20:15,550 code. And it's a curse in that, why 316 00:20:15,743 --> 00:20:19,281 hire junior programmers? And 317 00:20:20,591 --> 00:20:24,391 it's also a curse because people use it, I have seen 318 00:20:24,391 --> 00:20:27,158 it firsthand, people use it that don't know the material. 319 00:20:27,870 --> 00:20:31,303 And if you use it that way, you are in trouble. 320 00:20:32,081 --> 00:20:35,846 Because It is— if you have to understand 321 00:20:35,846 --> 00:20:38,818 how it was trained, it's trained off Reddit, and there are 322 00:20:38,818 --> 00:20:42,642 crazy ideas on Reddit, and there are some awesome 323 00:20:42,915 --> 00:20:46,465 answers in Reddit, and AI can't tell the difference. 324 00:20:47,043 --> 00:20:50,818 No. Right? It's— so if you get one of 325 00:20:50,818 --> 00:20:54,288 the crazy answers, you're gonna make a fatal flaw 326 00:20:54,594 --> 00:20:57,308 and a fatal mistake, right? So 327 00:20:58,288 --> 00:21:01,614 you gotta— you use it as a tool. I worry 328 00:21:01,726 --> 00:21:05,486 that we're eliminating all the jobs of the entry-level programmers and 329 00:21:05,567 --> 00:21:09,262 coders, and that's going to kill this industry down the 330 00:21:09,262 --> 00:21:12,974 road. I feel like we're poisoning the pipe. We're 331 00:21:12,974 --> 00:21:16,204 shutting off the pipe of new talent. I don't think that's going to end well 332 00:21:16,204 --> 00:21:20,012 for anyone. Nope. Nope. And the other thing 333 00:21:20,044 --> 00:21:23,884 is, AI can be— is great, and you can give it a prompt, 334 00:21:23,981 --> 00:21:27,802 and that's a whole nother science that you have to get to get, but 335 00:21:28,379 --> 00:21:31,858 at some point, that's got to move into production. 336 00:21:32,627 --> 00:21:36,073 And how do you know without thorough 337 00:21:36,073 --> 00:21:39,664 testing? And how do you test without the knowledge 338 00:21:39,776 --> 00:21:43,527 of what a test is and what the purpose is? I'm not saying you can't 339 00:21:43,544 --> 00:21:47,295 buy a tool that says, hey, it tests, or vibe 340 00:21:47,295 --> 00:21:51,078 coding will tell you, oh, we tested. What does that mean? The 341 00:21:51,398 --> 00:21:55,021 AI vibe coded tests. Right. Yeah. 342 00:21:55,470 --> 00:21:58,715 It spun up a page and it turned on the page and it said the 343 00:21:58,715 --> 00:22:02,432 page was rendered. Big whoop. How about you 344 00:22:02,464 --> 00:22:05,694 multiply that 4,000 times, like as though there's 345 00:22:05,694 --> 00:22:09,246 4,000 people logged in? Oh, we didn't do that type of 346 00:22:09,246 --> 00:22:12,700 test. We don't do that. Right? So 347 00:22:13,102 --> 00:22:16,853 we're missing that skill set that goes from 348 00:22:17,063 --> 00:22:20,766 quick to market to durable. So the 349 00:22:20,766 --> 00:22:24,592 future is going to have a whole lot of junk code in it. But, 350 00:22:25,346 --> 00:22:29,118 you know, it's— having been in this industry, uh, 351 00:22:29,359 --> 00:22:32,858 a long time, it's just another tool, and 352 00:22:32,954 --> 00:22:35,555 it's great. It's really a powerful tool if used right. 353 00:22:36,566 --> 00:22:40,145 And, um, it's done— I've done some wonderful 354 00:22:40,193 --> 00:22:43,259 designs with it, and it saved me a whole lot of time. 355 00:22:44,608 --> 00:22:48,299 It's, it's really efficient in ideation and making it part of, 356 00:22:48,364 --> 00:22:51,558 like, your brainstorming loop. Give me— today on a call I was on 357 00:22:52,232 --> 00:22:56,062 we were struggling to come up with a kind of one-word phrase, one sentence. 358 00:22:56,623 --> 00:23:00,244 Perfect. It's perfect for that. And I basically said, give me 15 359 00:23:00,244 --> 00:23:03,786 options, right? Right. And 360 00:23:04,090 --> 00:23:07,520 10 of them were— 10 of them were okay, right? 5 though, 361 00:23:08,337 --> 00:23:11,590 and one of them was just chef's kiss, right? And 362 00:23:12,503 --> 00:23:16,269 I think— But you knew the material you were looking for, Frank. That's my point. 363 00:23:16,654 --> 00:23:19,425 You knew when you saw the chef's kiss, 364 00:23:21,008 --> 00:23:24,386 you could appreciate it and it matched what your vibe was. 365 00:23:24,903 --> 00:23:28,647 Okay, now that's something esoteric. I'm talking about 366 00:23:28,840 --> 00:23:32,549 also, what if it's actual code, like processing 367 00:23:32,581 --> 00:23:36,277 code, and it's guessing or thinks it has it 368 00:23:36,277 --> 00:23:40,053 right and it ran through a series of tests? What happens 369 00:23:40,086 --> 00:23:41,919 when that needs to go full scale? 370 00:23:44,364 --> 00:23:48,099 Yeah, we don't know. I mean, no one— unless you have— unless you have— 371 00:23:48,211 --> 00:23:50,696 one of the things that, that I remember interviewing a guy 372 00:23:51,882 --> 00:23:55,729 and for a job, and I liked him, he was smart and all, and I 373 00:23:55,729 --> 00:23:59,497 was like, the only thing he doesn't have is the scars, right? 374 00:23:59,545 --> 00:24:03,360 Because sometimes what they say about good judgment is often the result 375 00:24:03,360 --> 00:24:07,047 of bad judgment. Yeah, yeah. Something you learn far more from your 376 00:24:07,047 --> 00:24:10,624 failures than you do your successes. And again, 377 00:24:11,283 --> 00:24:14,077 not going to name the guy, but he was very difficult to work with because 378 00:24:14,253 --> 00:24:17,722 it was kind of like he was very smart and he was always trying to 379 00:24:17,722 --> 00:24:21,512 prove that he was smart, but he didn't have the scars to be like, look, 380 00:24:21,576 --> 00:24:24,113 I know what the general best practice is, 381 00:24:25,189 --> 00:24:28,449 but sometimes you have to know— sometimes you have to be bopped on the nose 382 00:24:28,481 --> 00:24:31,082 and get a couple of bloody noses and realize, well— 383 00:24:31,275 --> 00:24:34,824 100%. Sometimes the 384 00:24:34,904 --> 00:24:38,676 best practice isn't worth going that doing that because it's going to cause 385 00:24:38,676 --> 00:24:42,270 these other problems downstream. Right. And like, you know, you want 386 00:24:42,270 --> 00:24:46,104 to— I think you're right, you know, having those scars, 387 00:24:46,201 --> 00:24:49,586 they're unpleasant experiences as you go through them. But, 388 00:24:49,730 --> 00:24:53,099 um, going through them, that's— it's quite 389 00:24:53,099 --> 00:24:56,821 unpleasant, but it's what makes you a better developer. And I 390 00:24:56,885 --> 00:25:00,302 think if we shut off the pipe, you know, when people our age 391 00:25:00,399 --> 00:25:02,677 decide— well, you're, you know, you and I are, you know, 392 00:25:05,068 --> 00:25:08,663 I'm 10 years younger than you. Exactly right. 393 00:25:08,711 --> 00:25:12,418 So, you know, in 20 years from now, you and I will both be sipping 394 00:25:12,851 --> 00:25:16,237 Mai Tais in a beach somewhere. You know, like— I will have a 395 00:25:16,301 --> 00:25:19,960 keyboard in my hand. I probably will have this tablet in my hand. 396 00:25:20,393 --> 00:25:24,165 But like, and I actually like what I do, right? So like, I've had 397 00:25:24,165 --> 00:25:26,893 this conversation with retirement planners when they come like, what do you want to do 398 00:25:26,893 --> 00:25:29,589 when you retire? And I'm like, I kind of like what I do, you know? 399 00:25:30,104 --> 00:25:33,589 But when we are no 400 00:25:33,589 --> 00:25:36,962 longer in the workforce, like we're cutting off, you know, the 401 00:25:37,123 --> 00:25:40,915 generation, maybe 2 generations below us, like they'll be the last 402 00:25:40,915 --> 00:25:44,642 people that know what it's like to run a DevOps thing and get their 403 00:25:44,771 --> 00:25:48,497 you-know-what handed to them because the deployment didn't go the way 404 00:25:48,804 --> 00:25:52,451 we thought it would. And, you know, you've been on those deployments where it works 405 00:25:52,451 --> 00:25:56,164 on my machine was an actual— people said that works on my machine. 406 00:25:56,261 --> 00:25:59,270 Yeah. Right? Now— Oh, I've had that happen. I've 407 00:25:59,270 --> 00:26:02,770 had that happen in scale 408 00:26:02,770 --> 00:26:05,970 here. I can't imagine what that would be like. This 409 00:26:05,970 --> 00:26:09,490 code worked on my laptop. Why isn't it running on the 410 00:26:09,490 --> 00:26:13,010 4,000 servers we have? Right, right, right, right, 411 00:26:13,010 --> 00:26:16,810 right, right. You know, I had a miniature version of that. 412 00:26:16,810 --> 00:26:20,610 I had an application that had like 200 users, and I'm still in school 413 00:26:20,610 --> 00:26:23,890 while I'm doing this. And like, you know, I pushed an update and I was 414 00:26:23,890 --> 00:26:26,570 like, it blew up on half the machines. And I'm like, it worked on my 415 00:26:26,570 --> 00:26:29,981 machine. And like, And then my 416 00:26:30,221 --> 00:26:33,046 manager at the time just shot me this look of like, 417 00:26:33,768 --> 00:26:37,460 like Jedi mind trick. Like I could just like, all right, I won't say that 418 00:26:37,460 --> 00:26:41,183 again. But, um, but no, that really taught me the, 419 00:26:41,312 --> 00:26:44,891 the, that, that little bloody nose taught me like really 420 00:26:45,228 --> 00:26:48,920 anticipate that the production, pushing it into production is gonna 421 00:26:48,920 --> 00:26:52,386 be a process that is gonna be unpleasant. Right. 422 00:26:53,093 --> 00:26:55,789 You know, or at least be prepared for it. Right. It's not gonna be as 423 00:26:55,789 --> 00:26:59,626 simple as compile and ship. I have learned as 424 00:26:59,626 --> 00:27:03,431 CEO far more from my failures 425 00:27:03,849 --> 00:27:07,285 than I have my successes. Now, we don't lose customers, 426 00:27:07,558 --> 00:27:11,171 right? Most MSPs, they did— the marketing group did 427 00:27:12,054 --> 00:27:15,361 what's called Net Promoter Score, and most 428 00:27:15,907 --> 00:27:19,616 technology companies are 30 to 35, and we're 92. 429 00:27:20,467 --> 00:27:24,216 Oh, wow. Yeah. So we don't lose too 430 00:27:24,216 --> 00:27:27,383 many customers. So we lost a customer 431 00:27:28,650 --> 00:27:31,752 4 years ago, and I 432 00:27:31,833 --> 00:27:34,480 was like, how did we lose them? 433 00:27:36,538 --> 00:27:40,219 And studying it and looking at each 434 00:27:40,509 --> 00:27:43,738 layer of where we failed in 435 00:27:44,626 --> 00:27:47,764 listening to them, in what their needs were, 436 00:27:48,503 --> 00:27:52,282 We were on autopilot. We were just doing a 437 00:27:52,460 --> 00:27:56,271 good day, average work on a good day, except 438 00:27:56,271 --> 00:27:59,772 the client was shifting and we completely 439 00:28:00,062 --> 00:28:03,434 missed the importance of the shift to the 440 00:28:03,434 --> 00:28:06,968 organization. Interesting. And 441 00:28:07,049 --> 00:28:09,787 that not feeling heard 442 00:28:11,423 --> 00:28:13,076 caused that client to go elsewhere. 443 00:28:15,460 --> 00:28:16,604 And that caused me 444 00:28:19,086 --> 00:28:22,619 first soul searching, right? And then you have to say, okay, 445 00:28:22,619 --> 00:28:25,864 that's the first 24 hours. It's the gut punch. 446 00:28:26,574 --> 00:28:30,412 And then it's how can you use this to improve? How 447 00:28:30,412 --> 00:28:33,358 can you use this to build a better product to 448 00:28:34,357 --> 00:28:38,044 keep going forward? And then you keep marching because the 449 00:28:38,125 --> 00:28:41,614 only thing you can change is your actions today. Right. You can't go in the 450 00:28:41,614 --> 00:28:45,367 past, and complaining about the past does absolutely 451 00:28:45,399 --> 00:28:49,200 nothing. There's no value because the present is the only 452 00:28:49,216 --> 00:28:51,974 place that you can be. So how do we be more 453 00:28:52,103 --> 00:28:55,856 present for our customers so that we can keep moving forward? 454 00:28:56,754 --> 00:29:00,475 What was your mitigation strategy? Do you just kind of touch base with more 455 00:29:01,646 --> 00:29:05,415 of your customers? Mitigation strategy? I, I hired new people 456 00:29:05,431 --> 00:29:09,248 to do things. We now have a VCIO that visits all of our 457 00:29:09,280 --> 00:29:12,956 clients. I divided the responsibilities across multiple 458 00:29:13,036 --> 00:29:16,585 people. My CTO is now the person who does the 459 00:29:16,585 --> 00:29:19,925 design. Every time we get 460 00:29:20,374 --> 00:29:24,131 especially larger accounts, he's the one who does the deep dive and 461 00:29:24,131 --> 00:29:26,990 does the initial design, and that's all he does. 462 00:29:28,065 --> 00:29:31,341 His designs are solid. Now there's a team of people 463 00:29:31,823 --> 00:29:35,467 at each layer, then the directors of the groups 464 00:29:35,548 --> 00:29:39,174 do the execution of the design. And if they run into a 465 00:29:39,174 --> 00:29:42,896 problem, they have one, one person, because what I didn't 466 00:29:42,896 --> 00:29:46,423 want to lose was the accountability. So it's— the 467 00:29:46,747 --> 00:29:50,465 goal is how do you divide up the work so that you don't lose the 468 00:29:50,465 --> 00:29:54,183 accountability? Because we had an account once, they had to get us 469 00:29:54,183 --> 00:29:56,496 into a firewall. There was 25 people on the call. 470 00:29:59,852 --> 00:30:03,548 Wow. What does that tell me? There's no accountability. And it 471 00:30:03,548 --> 00:30:07,078 tells, it tells me through one action how to get on the 472 00:30:07,078 --> 00:30:10,768 firewall. I, my, my director, one of the 473 00:30:10,768 --> 00:30:14,586 directors sent me the picture of the Teams meeting and there was literally 25, 474 00:30:14,778 --> 00:30:18,532 and he's, and he said, what do I do with this? And I said, first 475 00:30:18,532 --> 00:30:22,206 you have to cooperate with everybody and get logged in. But this 476 00:30:22,335 --> 00:30:25,335 is telling us this is the problem that this customer has. 477 00:30:25,993 --> 00:30:28,345 Every decision It's 25 people. 478 00:30:29,882 --> 00:30:33,554 Then if it's 25 people, there's no one throat to choke. I first 479 00:30:33,667 --> 00:30:37,516 heard— I, um, I first heard that term when 480 00:30:37,726 --> 00:30:41,089 I worked for a company and, uh, the customer was, um, 481 00:30:41,833 --> 00:30:45,551 the Marine Corps, actually. Yeah. And, um, I 482 00:30:45,551 --> 00:30:49,276 forget his rank or whatever, but fairly highly placed guy. And 483 00:30:49,292 --> 00:30:52,435 he goes— he was telling like, I want one throat to choke. 484 00:30:53,024 --> 00:30:56,597 My first thought was, well, that's his industry, so to speak. 485 00:30:56,677 --> 00:31:00,267 But that's a great, that's a great phrase, and I'm glad it's kind of caught 486 00:31:00,267 --> 00:31:03,936 on, right? Yeah, yeah. And it's, in a nutshell, very, 487 00:31:04,113 --> 00:31:07,894 very clear as to what is going on. And so 488 00:31:08,279 --> 00:31:11,965 here's how you, you massage that, because you can overwhelm that one 489 00:31:12,125 --> 00:31:15,602 throat, right? Yeah, absolutely. If you break the tasks 490 00:31:15,730 --> 00:31:18,967 up and each one has a throat, 491 00:31:19,740 --> 00:31:23,000 You can build a chain structure that all can move 492 00:31:23,000 --> 00:31:26,840 simultaneously. And that's the— that's, to me, that's 493 00:31:26,840 --> 00:31:30,625 the music. When you get there and you can watch an organization 494 00:31:30,625 --> 00:31:34,273 start to pivot and actually move in a direction, that's awesome. That 495 00:31:34,337 --> 00:31:36,233 feels good. That's cool. 496 00:31:38,904 --> 00:31:41,772 How do you balance 497 00:31:43,334 --> 00:31:47,129 autonomous operation of these teams, like where they can operate independently, 498 00:31:47,329 --> 00:31:49,102 with accountability? Is it breaking it down into small tasks, or is there a more 499 00:31:49,102 --> 00:31:52,811 nuanced It's that 500 00:31:52,811 --> 00:31:56,303 plus oversight. 501 00:31:56,529 --> 00:31:59,149 Oversight and accountability go hand in hand. 502 00:32:01,220 --> 00:32:04,701 We hold a— I 503 00:32:04,814 --> 00:32:08,659 coined the phrase, see the field. I'm the 504 00:32:08,676 --> 00:32:10,872 one— my CTO runs it. 505 00:32:12,900 --> 00:32:16,117 Everybody in our tier 4 in our organization is the top. 506 00:32:17,438 --> 00:32:20,616 level of engineers. We all get in a conference room 507 00:32:21,021 --> 00:32:24,376 twice a week, 11 o'clock Mondays and 11 o'clock Thursdays, 508 00:32:25,142 --> 00:32:27,599 and I'm a fly on the wall 509 00:32:29,507 --> 00:32:32,271 and I'm listening. Are we following process? 510 00:32:33,387 --> 00:32:36,933 Are we getting into— So I, I've got one client that we 511 00:32:36,997 --> 00:32:40,809 have worked for 3 months, 2 and a half months, to 512 00:32:40,874 --> 00:32:44,424 get the generator repaired by the electrician, 513 00:32:46,266 --> 00:32:49,560 and this data center is exposed because 514 00:32:50,079 --> 00:32:52,432 they're weak on power without this generator. 515 00:32:53,890 --> 00:32:57,690 And we've— the Tier 4 engineer who's been assigned 516 00:32:58,353 --> 00:33:02,040 has done that. And I said, stop. I said, I don't want to— and 517 00:33:02,315 --> 00:33:05,129 I literally stopped the meeting. I said, I don't want a Tier 4 engineer 518 00:33:05,912 --> 00:33:09,464 worrying about an electrician. That The customer 519 00:33:09,464 --> 00:33:13,273 needs to own the problem. It's their electrician, it's their 520 00:33:13,386 --> 00:33:16,373 generator. Get the 521 00:33:16,696 --> 00:33:20,217 customer, the main customer contact, and tell them 522 00:33:20,411 --> 00:33:23,964 this project is stuck until you clear that. 523 00:33:25,579 --> 00:33:29,131 So that's the one throat to choke, right? Is it's now, it's now in the 524 00:33:29,131 --> 00:33:32,655 customer's— and what you're doing, that action is empowering. So here, 525 00:33:33,371 --> 00:33:37,107 so oversight. Here's what I did in that one action. I 526 00:33:37,204 --> 00:33:40,845 empowered the client to help in their 527 00:33:40,893 --> 00:33:44,150 process. I empowered my Tier 4 engineer 528 00:33:44,599 --> 00:33:47,680 to ask for help and to understand where 529 00:33:48,690 --> 00:33:51,434 we— what is the limit of our service 530 00:33:52,525 --> 00:33:56,118 and how we can't do everything all the time. We have to 531 00:33:56,150 --> 00:33:58,557 keep our eye on the ball, which is solve this problem. 532 00:33:59,744 --> 00:34:02,405 And I'm doing it in a 533 00:34:03,286 --> 00:34:06,684 large meeting, so everybody is learning from it 534 00:34:06,956 --> 00:34:10,017 because I'm talking about something they have direct 535 00:34:10,978 --> 00:34:14,808 process. They're doing the actual work at that moment. So 536 00:34:14,808 --> 00:34:17,965 if you're doing work at that moment, that is the most 537 00:34:18,334 --> 00:34:22,132 influential time to make a change. If I sit here and I 538 00:34:22,132 --> 00:34:25,048 said, here's the 10 principles, follow these 10 principles, 539 00:34:25,465 --> 00:34:29,247 you'll Somebody will say, oh, those are great principles. I don't know how to apply 540 00:34:29,247 --> 00:34:32,868 them, so I'll forget them in 5 minutes. But if I say 541 00:34:33,589 --> 00:34:37,291 something right at that moment, do it this way, 542 00:34:37,499 --> 00:34:40,624 give this, give this issue to the client. 543 00:34:41,826 --> 00:34:45,159 Right. Now the engineer is saying, oh, I thought we'd take care of our clients. 544 00:34:45,351 --> 00:34:48,765 We are taking care of our clients by empowering them 545 00:34:49,213 --> 00:34:52,369 to help themselves and to help us. help them. 546 00:34:53,844 --> 00:34:57,387 And then it's like, ah, okay, I get it. So we do that 547 00:34:57,772 --> 00:35:01,298 twice a week. It's a 30-minute meeting and it's all high 548 00:35:01,443 --> 00:35:04,392 level. What are the obstacles you're facing? Where are you? And 549 00:35:04,969 --> 00:35:08,640 then I'll, I'll call audibles and I'll say, what about this? And 550 00:35:09,378 --> 00:35:13,097 it's— my CTO runs it. I do color commentary. 551 00:35:14,267 --> 00:35:18,018 And because I want him to own the problem and 552 00:35:18,372 --> 00:35:22,089 That's the model. So I model for my staff how I 553 00:35:22,137 --> 00:35:25,823 want them to manage their people. And we have an 554 00:35:25,903 --> 00:35:29,716 overseas division too. Our tier 1 and 2 is 555 00:35:29,973 --> 00:35:33,803 overseas, and people don't leave that company 556 00:35:33,979 --> 00:35:37,184 either. Generally, people don't leave Harbor. 557 00:35:37,744 --> 00:35:41,478 They stay. I have people for 13, 15 years. Oh, wow. 558 00:35:42,327 --> 00:35:45,099 And it's not money. Money is important, 559 00:35:48,291 --> 00:35:51,942 It's culture. Culture is critical in an organization. And 560 00:35:52,283 --> 00:35:56,083 it's the best thing about what I do and working in technology 561 00:35:57,321 --> 00:36:00,949 is the people. I get a chance to grow people, right? So I 562 00:36:00,998 --> 00:36:04,756 had the manager, we have one accountable manager overseas. 563 00:36:07,779 --> 00:36:11,541 And I said to him, how come there's jobs everywhere? How come people 564 00:36:11,541 --> 00:36:14,400 stay here? He said, because it's the way you— 565 00:36:16,061 --> 00:36:19,825 this organization operates. And I said, well, give me an example. He said, 566 00:36:20,814 --> 00:36:24,641 when you want to talk to somebody, you never say— you never just 567 00:36:24,641 --> 00:36:27,789 call them and say, hey, 568 00:36:29,115 --> 00:36:32,687 do this. You always say, do you have a moment? 569 00:36:34,772 --> 00:36:36,920 He said that little thing means 570 00:36:38,604 --> 00:36:42,064 Respect their time. That's right. And so when the 571 00:36:42,064 --> 00:36:45,493 CEO is doing that, then 572 00:36:45,862 --> 00:36:49,670 the chiefs are doing it, and then the directors are doing it. 573 00:36:50,570 --> 00:36:53,959 And it's what he's telling me, what he's telegraphing to me, is it 574 00:36:53,959 --> 00:36:57,624 percolated all the way down to him, which 575 00:36:57,656 --> 00:37:01,321 says what I'm really care— what I'm really doing is making 576 00:37:01,353 --> 00:37:04,459 sure, does everybody Are we following our 577 00:37:04,459 --> 00:37:08,139 corporate values? Is everybody on board? 578 00:37:08,894 --> 00:37:11,690 The other thing, to digress, the other thing those meetings do 579 00:37:12,574 --> 00:37:15,756 is educate, because what we miss 580 00:37:16,511 --> 00:37:19,982 in the world we have is, again, the information silos are 581 00:37:19,982 --> 00:37:23,806 vertical. Right. So I'm not worried, I'm not worried, will my guys 582 00:37:23,919 --> 00:37:27,100 get the vertical education? I worry about the horizontal education. 583 00:37:29,270 --> 00:37:32,971 That's the opportunities for, for me to, to dive 584 00:37:32,971 --> 00:37:36,529 into the horizontal education and talk about 585 00:37:37,058 --> 00:37:39,974 people skills and how to deal with certain people and how to 586 00:37:40,583 --> 00:37:44,253 overcome obstacles. And, um, 587 00:37:45,327 --> 00:37:48,948 that's what I love my job. I mean, I love it. You can tell, you 588 00:37:48,948 --> 00:37:52,105 can tell you have a passion for this. And, you know, yeah, if you're doing 589 00:37:52,137 --> 00:37:54,958 something for decades, like in tech, like 590 00:37:56,532 --> 00:38:00,273 You have to love it, right? Oh, you have to. Yeah, yeah. No, I tell 591 00:38:00,321 --> 00:38:04,094 people, if you don't— I— people, oh, I really need a 592 00:38:04,094 --> 00:38:07,851 job. No, you need a passion. I— because the first 593 00:38:07,851 --> 00:38:11,464 4 clients that I got were all barter clients. Why? Because I 594 00:38:11,464 --> 00:38:15,286 needed somebody to teach me, right? When it was just 595 00:38:15,398 --> 00:38:19,107 me, I needed somebody to help teach me how 596 00:38:19,107 --> 00:38:22,459 to do my job. I knew how to solve the tech problems. 597 00:38:22,814 --> 00:38:26,577 That wasn't my job. My job was to solve their problems. 598 00:38:27,680 --> 00:38:31,187 And so those first 4 clients, 599 00:38:32,222 --> 00:38:35,716 3 of which— 2 of which I still have today. 600 00:38:36,764 --> 00:38:40,153 Wow. Yeah. 30 years later. 601 00:38:42,920 --> 00:38:46,643 Wow. So clearly you're a successful entrepreneur. Yeah, no, 602 00:38:46,723 --> 00:38:50,215 I'm— yeah, no, I, I have 70 603 00:38:50,327 --> 00:38:53,243 employees. Wow. Yeah. You know, I— 604 00:38:54,012 --> 00:38:57,665 my CTO, my director, director of technology, 605 00:38:57,938 --> 00:39:01,735 my COO, and I go to India once a year. Why? 606 00:39:02,087 --> 00:39:05,884 We don't— does anybody ever go? We all outsource, right? 607 00:39:06,349 --> 00:39:09,473 Very few people do. Yeah, I go. 608 00:39:10,178 --> 00:39:13,607 Why? Because, A, those are my people. I opened an Indian 609 00:39:13,703 --> 00:39:17,257 subsidiary. I didn't farm out to some third 610 00:39:17,289 --> 00:39:20,728 party where I have absolutely no control or input. 611 00:39:21,452 --> 00:39:25,104 I created an organization there that matches the culture 612 00:39:25,639 --> 00:39:29,400 of the organization here. And by 613 00:39:29,464 --> 00:39:32,398 doing that, I can extend the quality and take 614 00:39:32,883 --> 00:39:35,675 care— take better care of my clients. Right. Interesting. 615 00:39:41,497 --> 00:39:45,276 So you mean you, you have instilled that culture at every 616 00:39:45,292 --> 00:39:48,163 level? It sounds— I have to. Yeah, I have to. 617 00:39:49,452 --> 00:39:52,954 Have to. And you have to let 618 00:39:52,987 --> 00:39:56,778 people fail because, like I said, you only learn from your 619 00:39:56,778 --> 00:39:59,228 failures. And then I got to pick them up. 620 00:40:01,014 --> 00:40:04,725 And when they fail, okay, what did you— what— 621 00:40:05,209 --> 00:40:08,732 again, when I lost a customer, I failed. I failed. 622 00:40:08,957 --> 00:40:12,259 My company failed. Okay, deal with the gut punch. 623 00:40:13,548 --> 00:40:16,672 Be paralyzed for 24 hours. Now come back. 624 00:40:17,560 --> 00:40:20,415 What do we need to do? How do we fix this? 625 00:40:21,760 --> 00:40:25,452 So as someone who's been in the technology game for as long as you 626 00:40:25,452 --> 00:40:28,741 have, and you've seen all these changes, what still separates 627 00:40:28,741 --> 00:40:31,888 organizations that build durable systems 628 00:40:33,518 --> 00:40:37,163 from those that just spend their time with punching tickets and like putting 629 00:40:37,232 --> 00:40:41,006 fires out? Their 630 00:40:41,070 --> 00:40:44,595 value prop internally, their value— Culture, kind of? 631 00:40:45,317 --> 00:40:48,233 Yeah, their culture, their core values. Do they 632 00:40:49,098 --> 00:40:52,367 promote and people would promote and 633 00:40:52,447 --> 00:40:56,277 enhance people with passion, or do they demote 634 00:40:56,630 --> 00:41:00,379 people with passion? They scare people. And 635 00:41:00,460 --> 00:41:04,049 passion can be two— it's a curse and it's a blessing. 636 00:41:05,429 --> 00:41:08,574 Just like AI. It's like a lot of things, right? 637 00:41:09,136 --> 00:41:12,731 Just like a lot of things. Food. I mean, I used to be 300 638 00:41:12,731 --> 00:41:16,309 pounds, right? You know, I loved— I used to eat a pizza all 639 00:41:16,502 --> 00:41:20,241 by myself. It was killing me. And if I wanted to be here longer, I 640 00:41:20,289 --> 00:41:23,964 had to make changes. So I had to, you know, 641 00:41:24,061 --> 00:41:27,399 why did I eat like that? I had to deal with that. 642 00:41:28,378 --> 00:41:31,781 Now it's a lot easier. I mean, GLP-1s are amazing. And yeah, 643 00:41:32,809 --> 00:41:36,489 you know, they, uh, but I, I was doing the, 644 00:41:36,650 --> 00:41:40,201 the precursors, growth hormones, before that because I got the 645 00:41:40,330 --> 00:41:44,174 DNA of an obese person. And if I, if I 646 00:41:44,286 --> 00:41:46,329 don't— so 647 00:41:48,243 --> 00:41:52,080 I don't remember the tangent I was on, but— What 648 00:41:52,080 --> 00:41:54,811 separates companies that actually build durable systems and, 649 00:41:55,803 --> 00:41:58,773 and versus ones that just They're always putting fires 650 00:41:58,886 --> 00:42:02,672 out. Putting fires out tells me 651 00:42:02,977 --> 00:42:06,602 right off the top that they've not tested enough. Right. 652 00:42:07,292 --> 00:42:10,902 And they haven't thought about failure, which means they're 653 00:42:10,998 --> 00:42:14,656 immature, which means they're not really 654 00:42:14,833 --> 00:42:18,394 paying homage to wanting to be 655 00:42:19,100 --> 00:42:22,871 okay. Okay, that's one way to be. 656 00:42:24,298 --> 00:42:27,925 It's not the way to be successful in my book, and it's 657 00:42:27,925 --> 00:42:31,631 not the way to be long-term successful. Right. So 658 00:42:32,321 --> 00:42:35,498 any process, you don't say, 659 00:42:36,734 --> 00:42:40,312 oh, this is perfect. You say, well, what if that breaks? What if that 660 00:42:40,440 --> 00:42:43,023 breaks? What if that breaks? What do we do? 661 00:42:44,195 --> 00:42:47,709 And you can get to the laws of diminishing 662 00:42:47,805 --> 00:42:51,319 return where if you've got so many things, you're like, okay, this is good. 663 00:42:51,977 --> 00:42:55,408 Now we can put it in production and it will teach us 664 00:42:58,113 --> 00:43:01,797 what we need to know. Right. I mean, 665 00:43:02,412 --> 00:43:06,068 I, uh, we had a, a unit from a hardware 666 00:43:06,133 --> 00:43:09,798 vendor, big storage unit. It was 667 00:43:09,895 --> 00:43:13,513 up for— we— my client 668 00:43:13,529 --> 00:43:16,259 bought it in March. It's still not in production. 669 00:43:18,313 --> 00:43:20,453 They deployed it. It was a storage cluster. 670 00:43:22,262 --> 00:43:26,079 And I don't want to say the vendor because I don't— I'm not interested in 671 00:43:26,675 --> 00:43:27,593 badmouthing people. 672 00:43:31,204 --> 00:43:34,950 But we were on— my guys were 673 00:43:34,999 --> 00:43:38,277 on umpteen phone calls and meetings 674 00:43:39,359 --> 00:43:42,477 to get this damn thing built. And they came out with their 675 00:43:42,526 --> 00:43:44,768 engineers and they set it up. 676 00:43:45,787 --> 00:43:49,531 I'm at the point of, could we just buy 4 677 00:43:49,659 --> 00:43:53,098 pieces of equipment and put drives and build our own cluster? Because 678 00:43:54,062 --> 00:43:57,453 this is never going to work. No, let's get— we have to work with the 679 00:43:57,453 --> 00:43:58,117 vendor. Okay. 680 00:44:01,498 --> 00:44:05,074 They built it. They said it's stable and it's good. You can use it. We 681 00:44:05,074 --> 00:44:08,860 used it. We then said internally, we 682 00:44:08,860 --> 00:44:12,597 don't use anything until we run it for 7 days and bang the crap 683 00:44:12,597 --> 00:44:16,369 out of it. Whatever it is, sure enough, day 684 00:44:16,450 --> 00:44:18,199 two, it crashes. 685 00:44:20,996 --> 00:44:24,771 Okay. Now, if we would have taken the 686 00:44:24,771 --> 00:44:28,305 vendor's word for it, and we could have, right? We've been told from the 687 00:44:28,321 --> 00:44:32,022 vendor this is going to be good. It's still it will 688 00:44:32,022 --> 00:44:35,476 fully work, and we've tested it, and 689 00:44:35,476 --> 00:44:39,060 it's it's a company that everybody knows. This is not I'm not 690 00:44:39,060 --> 00:44:42,428 talking about some. Overseas 691 00:44:42,428 --> 00:44:45,442 fly-by-night organization. I'm talking about top-line 692 00:44:46,038 --> 00:44:49,486 enterprise hardware. This is like a legit enterprise 693 00:44:49,583 --> 00:44:53,134 company, not somebody selling it from the back of a U-Haul truck, right? Yeah, no, 694 00:44:53,150 --> 00:44:56,707 no, no, no, no, no, no, this is with teams and 695 00:44:56,707 --> 00:44:58,615 teams of people, okay? Right, right, right. 696 00:45:00,999 --> 00:45:04,563 I could have, as the vendor, right? It's a vendor in my client's 697 00:45:04,563 --> 00:45:07,999 environment. We asked for storage, they ordered the storage. 698 00:45:10,157 --> 00:45:12,201 Their vendor is telling us it's all good. 699 00:45:14,711 --> 00:45:17,251 I— there's absolutely no reason that I shouldn't be using it, 700 00:45:18,335 --> 00:45:21,572 except I know history. 701 00:45:22,624 --> 00:45:26,182 So we have a process, we 702 00:45:26,182 --> 00:45:29,981 test, and we don't test a little bit, we 703 00:45:29,981 --> 00:45:33,393 throw a whole script at it. We're gonna 704 00:45:33,603 --> 00:45:36,748 knock this thing on its, on its ass and see how it performs. 705 00:45:38,023 --> 00:45:41,592 Why? Because when we put 300 VMs on there, 706 00:45:42,610 --> 00:45:45,937 guess what's going to happen? It's going to be on its ass. So we'd like 707 00:45:45,937 --> 00:45:49,640 to know what's going to happen now rather than later. Well, 708 00:45:49,657 --> 00:45:53,137 it's better to know before it hits the storm. You know, this 709 00:45:53,218 --> 00:45:56,682 reminds me of a story, and this is one of the worst 710 00:45:57,540 --> 00:46:01,043 scars I have. I was 711 00:46:01,301 --> 00:46:04,915 working at a moderate-sized consulting company, and 712 00:46:05,173 --> 00:46:08,938 when I moved to Richmond. And one of the projects they had, they 713 00:46:08,938 --> 00:46:12,774 did this joint project with a 714 00:46:12,823 --> 00:46:16,185 local university. And we had 715 00:46:16,201 --> 00:46:18,589 purchased software from— 716 00:46:20,880 --> 00:46:24,381 it was a guy who was an entrepreneur who had built a beverage company. Right. 717 00:46:24,445 --> 00:46:27,230 And he built like a survey software to kind of go with it, 718 00:46:28,309 --> 00:46:31,239 like to basically do his own customer thing. And then he spun that off. I 719 00:46:31,271 --> 00:46:34,056 think he sold the beverage company and then spun that off to its own thing. 720 00:46:34,842 --> 00:46:38,599 And, you know, we grilled them with questions and the 721 00:46:38,664 --> 00:46:42,357 budget was so tight that they decided to cut the testing cycle. 722 00:46:44,625 --> 00:46:48,471 Right. And we took the vendor's word for it of how 723 00:46:48,536 --> 00:46:52,066 it can scale. Yeah. Yeah. 724 00:46:52,874 --> 00:46:56,075 Yeah. Now, as I described that, 725 00:46:56,981 --> 00:47:00,462 you can see, you know, the iceberg sailing straight into the 726 00:47:00,462 --> 00:47:04,082 Titanic. Oh yeah. Yeah. But when we— when it— I mean, we 727 00:47:04,162 --> 00:47:07,697 bought new servers, but— and when the servers keeled over on the 728 00:47:07,713 --> 00:47:10,519 first day, right, what the hell's going on? 729 00:47:11,663 --> 00:47:15,096 Anyway, we did some like initial kind of triaging and we found out that 730 00:47:15,225 --> 00:47:16,561 like it would basically— 731 00:47:19,556 --> 00:47:23,259 it would basically not run any kind of database 732 00:47:23,259 --> 00:47:26,681 queries. It would basically pull back everything and then the data was 733 00:47:26,681 --> 00:47:30,126 sorted from in code as 734 00:47:30,142 --> 00:47:33,943 opposed to having the database do it, which right there, there's a massive 735 00:47:33,991 --> 00:47:37,263 bottleneck, right? Got it. I mean, so— 736 00:47:37,696 --> 00:47:41,432 The indexing was code-based and there wasn't a true index. There was 737 00:47:41,529 --> 00:47:45,073 no true index. So we're picking this apart and we're like, oh my 738 00:47:45,073 --> 00:47:48,906 God. And that's when I learned if you can't afford to test 739 00:47:48,906 --> 00:47:52,514 it, you can't afford to do it, right? Testing is not 740 00:47:52,514 --> 00:47:56,200 optional. And Andy Leonard has a, um, has a 741 00:47:56,200 --> 00:47:59,642 phrase, my co-host on the show who's normally here. He goes, all software is 742 00:47:59,674 --> 00:48:03,326 tested, uh, but sometimes the first time it's 743 00:48:03,326 --> 00:48:06,766 tested is in production. You know, 744 00:48:06,912 --> 00:48:10,595 and, and then those never ends well. 745 00:48:10,724 --> 00:48:14,456 And that's the type of like judgment call. Yeah. Anytime, like ever since then, 746 00:48:14,504 --> 00:48:18,055 this was like 2006. I'm still like, I 747 00:48:18,119 --> 00:48:20,915 cringe when I think about the, the consequences of that. 748 00:48:21,834 --> 00:48:25,590 So now, now cycle back to your question 749 00:48:25,590 --> 00:48:29,217 about AI. Yeah. Somebody vibe codes and 750 00:48:29,217 --> 00:48:32,877 says, oh, I want to scale this. Guess what they're going 751 00:48:33,021 --> 00:48:36,777 to do? The same hairs in the back of my neck. I get flashbacks. 752 00:48:37,323 --> 00:48:40,324 No, you're right. Like, they don't know. And, and, you know, it's very easy to 753 00:48:40,356 --> 00:48:44,096 say, look, we'll just throw more cloud resources at it. Yeah, you can 754 00:48:44,144 --> 00:48:47,482 do it. And I, I thank God that, like, that 755 00:48:47,498 --> 00:48:51,191 happened pre-cloud because it just keeled over 756 00:48:51,191 --> 00:48:53,904 like these 4 new, like brand new top-of-the-line servers. 757 00:48:55,044 --> 00:48:58,673 If you had that in the cloud, the bill would have been astronomical because the 758 00:48:58,978 --> 00:49:01,644 cloud just would have thrown more stuff at it, right? It's kind of like you 759 00:49:01,644 --> 00:49:05,064 see that video of the guy doing with money. That's music to my ears, 760 00:49:06,123 --> 00:49:09,816 right? That's my business. There's my business model. There's my business model. I 761 00:49:09,816 --> 00:49:13,108 can give you DevOps, SaaS 762 00:49:13,108 --> 00:49:16,222 infrastructure engineering, for 30% 763 00:49:16,720 --> 00:49:20,556 less than they're charging and actually have eyes 764 00:49:20,652 --> 00:49:24,199 on it, not just raw anything, so that the next 765 00:49:24,199 --> 00:49:27,938 time your code goes nuts, we'll stop it as we're seeing 766 00:49:28,131 --> 00:49:31,886 the problem and hook your engineer in the place and 767 00:49:31,902 --> 00:49:35,481 say, hey, come on, we need help. This— look what's happening. Right, right, right. And 768 00:49:35,497 --> 00:49:38,658 that's a better— even if you page out the engineer at 3 in the morning, 769 00:49:38,707 --> 00:49:41,579 that's a better— That's a better conversation 770 00:49:42,948 --> 00:49:46,593 Anytime. Than a $2 million bill at the end of the month. Yeah. Yeah. 771 00:49:46,658 --> 00:49:50,390 Yeah. Well, awesome. I see we're at time. I really 772 00:49:50,390 --> 00:49:54,129 enjoyed this conversation. We'd love to have you back. Sure. It's been really fun. Where 773 00:49:54,129 --> 00:49:57,928 can folks find out more about you and what you're up to? Email 774 00:49:57,960 --> 00:50:00,396 me, richardluna@protectedharbor.com. 775 00:50:02,680 --> 00:50:06,479 AI isn't waiting for your security to catch up, and neither 776 00:50:06,479 --> 00:50:09,559 should you. Thanks again to Richard Luna for joining us. 777 00:50:09,997 --> 00:50:12,869 Subscribe to Data Driven, and we'll see you next time.