1 00:00:00,040 --> 00:00:03,960 A hallucination can kind of be thought of as a, as a statement 2 00:00:03,960 --> 00:00:07,240 that's statistically probable, but factually incorrect. 3 00:00:07,240 --> 00:00:09,680 I, I, I like that. I read that statement somewhere. 4 00:00:09,680 --> 00:00:11,040 I was like, that's perfect way to describe it. 5 00:00:11,040 --> 00:00:13,520 Statistically probable, factually incorrect, right. 6 00:00:13,720 --> 00:00:16,560 And again, this is one of the reasons that people are building 7 00:00:17,040 --> 00:00:20,880 grounding systems around AIS like like RAG, because it's a 8 00:00:20,880 --> 00:00:23,400 safety mechanism, It's a security mechanism, like you 9 00:00:23,400 --> 00:00:25,280 said, right? To make it more, make the answer 10 00:00:25,280 --> 00:00:27,640 more reliable, make it more accurate. 11 00:00:28,600 --> 00:00:31,440 It's certainly a concern for us as we were, as we were building 12 00:00:31,440 --> 00:00:33,320 the product, building it into our product. 13 00:00:33,320 --> 00:00:37,000 And it's also why you see disclaimers like on chatbot 14 00:00:37,000 --> 00:00:38,880 systems. It's like, hey, LLMS can make 15 00:00:38,880 --> 00:00:40,680 mistakes. Please, please verify your 16 00:00:40,680 --> 00:00:43,600 answers because hallucinations are and will continue to be a 17 00:00:43,600 --> 00:00:48,040 problem even with grounding systems around around, you know, 18 00:00:48,040 --> 00:00:49,640 the, the implementation of these. 19 00:00:49,960 --> 00:00:52,480 So yeah, I think it's, I think it's a great question. 20 00:00:52,680 --> 00:00:56,440 Are hallucinations the the biggest threat to the success of 21 00:00:56,440 --> 00:00:59,560 an AI right now and trying to put guard reels around it so 22 00:00:59,560 --> 00:01:01,520 that you're not giving incorrect information? 23 00:01:01,520 --> 00:01:03,040 Or is there something else that keeps you up at night? 24 00:01:08,720 --> 00:01:13,800 This is identity at the center if it has anything to do with 25 00:01:13,840 --> 00:01:18,440 IAM. This is the go to podcast now 26 00:01:18,440 --> 00:01:22,320 your hosts Jim McDonald and Jeff Stedman. 27 00:01:28,320 --> 00:01:30,040 Welcome to the Identity of the Center podcast. 28 00:01:30,280 --> 00:01:31,760 I'm Jeff, and that's Jim. Hey, Jim. 29 00:01:32,160 --> 00:01:34,480 Hey, Jeff, how are you? Not so bad yourself. 30 00:01:35,160 --> 00:01:37,080 I just living the allergy life man. 31 00:01:37,080 --> 00:01:39,800 Even though we've we've gotten to the point of daylight savings 32 00:01:39,800 --> 00:01:43,960 time is over and temperatures are getting cooler, my allergies 33 00:01:43,960 --> 00:01:47,840 just keep getting worse. So if it sounds like I only 34 00:01:47,840 --> 00:01:49,520 breathe through my mouth, that's why. 35 00:01:50,600 --> 00:01:52,720 Well, yeah, if you could just not breathe and make sure the 36 00:01:52,720 --> 00:01:54,960 audio quality sounds good, that would be fantastic for me. 37 00:01:55,120 --> 00:01:57,440 Yeah sure I'll I'll die mid episode. 38 00:01:58,360 --> 00:02:00,800 Well, today's special episode because we've got a sponsored 39 00:02:00,800 --> 00:02:04,560 spotlight and we're joined by our friends from Strabacity. 40 00:02:05,400 --> 00:02:07,720 His name is Steven Cox, he's a Co founder and CTO. 41 00:02:07,720 --> 00:02:10,280 Well, welcome here in a second, but just to make it crystal 42 00:02:10,280 --> 00:02:13,360 clear, this entire episode is sponsored thanks to Stromacity. 43 00:02:14,400 --> 00:02:17,960 We work with folks like Steven all the time to try and come up 44 00:02:17,960 --> 00:02:19,680 with different entertaining shows around. 45 00:02:19,720 --> 00:02:21,400 You know, there are specific viewpoints into the different 46 00:02:21,400 --> 00:02:22,800 things that they approach from identity space. 47 00:02:22,800 --> 00:02:25,560 So I'm sure people enjoy it. We've had actually Steven on 48 00:02:25,560 --> 00:02:28,240 before in the show. Steven, you joined us way back 49 00:02:28,240 --> 00:02:31,280 in episode 112. So welcome back to the Identity 50 00:02:31,280 --> 00:02:33,400 of the Center podcast. Thanks a lot guys. 51 00:02:33,400 --> 00:02:34,680 Yeah, it's great. It's great to be on. 52 00:02:34,680 --> 00:02:37,840 Always enjoy talking with you guys and, and I'm proud to have 53 00:02:37,840 --> 00:02:41,400 been #112 on your on your podcast. 54 00:02:41,960 --> 00:02:43,760 An early supporter, we appreciate it. 55 00:02:44,520 --> 00:02:46,800 You were with Trivacity then, you're still with Trivacity now, 56 00:02:46,800 --> 00:02:49,480 and that website is Trivacity dot AI, so make sure you get it 57 00:02:49,480 --> 00:02:53,320 out there. STRIVACITY dot AI definitely 58 00:02:53,320 --> 00:02:55,600 encourage to be able to go check it out. 59 00:02:55,600 --> 00:02:58,560 But why don't we start sort of at the very beginning, you know, 60 00:02:58,560 --> 00:03:02,280 tell us about Trivacity and give us a refresher on how you got 61 00:03:02,280 --> 00:03:05,000 into the digital identity space. Yeah. 62 00:03:05,000 --> 00:03:08,760 So I've been in the digital identity space now for, gosh, 63 00:03:08,760 --> 00:03:12,800 about 10 years. I came from a sort of network 64 00:03:12,800 --> 00:03:14,840 and endpoint threat detection background. 65 00:03:15,840 --> 00:03:17,520 I worked in that space for a number of years. 66 00:03:17,520 --> 00:03:21,360 I've been in cybersecurity for probably about 20 years now. 67 00:03:22,880 --> 00:03:27,720 I joined about 10 years ago, I joined a company that your your 68 00:03:28,240 --> 00:03:32,400 listeners are probably familiar with secure auth with who is now 69 00:03:32,400 --> 00:03:34,320 my Co founder of Trivacity, Keith Graham. 70 00:03:35,240 --> 00:03:38,280 And so I focused on, you know, workforce identity management 71 00:03:38,280 --> 00:03:41,120 and customer identity management at at secure auth, which, you 72 00:03:41,120 --> 00:03:45,160 know, informed a lot of my, you know, my current beliefs around 73 00:03:45,800 --> 00:03:49,120 customer identity platforms. So, yeah, that's kind of how I 74 00:03:49,120 --> 00:03:54,240 got my start on, you know, in, in, in this space. 75 00:03:54,240 --> 00:03:58,920 And you know, Stervacity is a, is a customer IAM solution. 76 00:03:59,320 --> 00:04:03,440 We got started in about 2019. So gosh, we've been going about 77 00:04:03,440 --> 00:04:08,000 5-5 years, 5 1/2 years now. So we're focused on the, the 78 00:04:08,000 --> 00:04:10,520 customer identity problem end to end, right. 79 00:04:10,840 --> 00:04:13,080 You hear lots of, you know, lots of terms thrown around this 80 00:04:13,080 --> 00:04:15,680 space, customer identity, consumer identity, but 81 00:04:15,680 --> 00:04:18,959 fundamentally like we're ACCIM solution and that's the problem 82 00:04:18,959 --> 00:04:21,760 we're we're interested in solving for our customers. 83 00:04:22,840 --> 00:04:26,760 So what is it about your guys's approach to the consumer or 84 00:04:26,760 --> 00:04:28,520 customer because we're talking Siam, right? 85 00:04:28,520 --> 00:04:31,320 CIA am here. What makes you guys unique 86 00:04:31,320 --> 00:04:33,440 because I think that's the question that I always have for 87 00:04:33,440 --> 00:04:35,360 for folks in the space. OK, well, you know, there's lots 88 00:04:35,360 --> 00:04:37,440 of solution in the space. How are you guys different and 89 00:04:37,680 --> 00:04:41,440 how do your customers measure success with the, you know with 90 00:04:41,440 --> 00:04:43,240 your products? Yeah. 91 00:04:43,240 --> 00:04:46,920 So there's a couple things I focus on when we talk about 92 00:04:46,920 --> 00:04:49,440 differentiation, You know what, what sort of, you know, brings 93 00:04:49,440 --> 00:04:52,400 us out from the pack and and the and the three things I like to 94 00:04:52,400 --> 00:04:56,040 really talk about are are architecture, our capabilities 95 00:04:56,280 --> 00:04:58,960 and our focus on on customer experience. 96 00:04:58,960 --> 00:05:02,280 So in terms of architecture, you know, we spent a lot of our 97 00:05:02,280 --> 00:05:05,960 early days focused on building a platform that could scale with 98 00:05:05,960 --> 00:05:08,360 the unique needs of customer identity. 99 00:05:08,960 --> 00:05:12,200 You know, as we are building, you know, we kind of set some, 100 00:05:12,240 --> 00:05:15,600 some design principles out to, to help guide us, you know, and 101 00:05:15,600 --> 00:05:17,840 those principles of have evolved over the years, of course. 102 00:05:17,840 --> 00:05:19,800 But you know, I can give you a high level kind of what we were 103 00:05:19,800 --> 00:05:22,320 thinking. You know, one of the, one of the 104 00:05:22,320 --> 00:05:26,600 big ones was like going after the cloud hesitancy problem, 105 00:05:26,600 --> 00:05:28,360 right? They're, they're, you know, this 106 00:05:28,360 --> 00:05:30,680 is less of a problem than it was now than it, than it was in 107 00:05:30,680 --> 00:05:34,080 2019, but it is still an issue. And it centres around that 108 00:05:34,440 --> 00:05:38,240 hesitancy of organizations to move certain types of data or 109 00:05:38,240 --> 00:05:40,920 systems to the cloud for a variety of reasons, right. 110 00:05:41,160 --> 00:05:44,240 We learned a lot about this in previous experiences before 111 00:05:44,240 --> 00:05:46,520 servacity. So we wanted to really attack 112 00:05:46,520 --> 00:05:49,840 that head on. Another thing we sort of 113 00:05:50,200 --> 00:05:53,720 observed previous experiences was that CIM projects were sort 114 00:05:53,720 --> 00:05:57,480 of excruciatingly long and laborious, you know, kind of 115 00:05:57,480 --> 00:06:00,560 often for the wrong reasons. And we wanted to sort of go 116 00:06:00,560 --> 00:06:03,840 after that problem. You know, customer identity has 117 00:06:04,080 --> 00:06:07,240 some unique requirements. It requires you to to move 118 00:06:07,240 --> 00:06:11,280 quickly, to iterate quickly as sort of customer patterns change 119 00:06:11,840 --> 00:06:13,760 and you discover new customer patterns. 120 00:06:14,560 --> 00:06:18,000 We wanted to bring the whole customer identity journey under 121 00:06:18,040 --> 00:06:19,880 the umbrella of a single platform, right? 122 00:06:19,880 --> 00:06:23,200 Something that had not yet been achieved in 2019 and arguably 123 00:06:23,200 --> 00:06:27,480 still has not been achieved or only been achieved by us, by us 124 00:06:27,480 --> 00:06:30,920 today. And you know, sort of sort of 125 00:06:30,920 --> 00:06:34,160 goes along with that are are things like brand management, 126 00:06:34,160 --> 00:06:36,880 which was sort of horribly broken until we came along. 127 00:06:36,880 --> 00:06:39,000 So you have these, you know, you have these orgs that are sort of 128 00:06:39,000 --> 00:06:41,840 brand of brand organizations, you know, have a sort of a 129 00:06:41,840 --> 00:06:44,000 parent company and multiple sub brands. 130 00:06:44,680 --> 00:06:48,160 That's a challenge we run into. So yeah, so we developed an 131 00:06:48,160 --> 00:06:50,520 architecture that takes these challenge into into account, 132 00:06:50,520 --> 00:06:52,040 right? We're built on Kubernetes. 133 00:06:52,560 --> 00:06:56,120 Our platform scales with demand. It can be upgraded using rolling 134 00:06:56,120 --> 00:06:58,120 upgrades, which means no downtime. 135 00:06:59,200 --> 00:07:02,080 You know, each of our customers, and this is a big point, gets an 136 00:07:02,080 --> 00:07:04,200 isolated instance of our product, right? 137 00:07:04,200 --> 00:07:08,880 We call this isolation by design and that's really what we're 138 00:07:08,880 --> 00:07:12,040 what we're doing to sort of address that cloud hesitancy 139 00:07:12,040 --> 00:07:14,240 issue, right. It also helps with sort of data 140 00:07:14,240 --> 00:07:16,800 sovereignty requirements. But we believe this is sort of 141 00:07:16,800 --> 00:07:19,880 superior to multi tenant architectures. 142 00:07:20,880 --> 00:07:23,080 And then orchestration, we have an intrinsic orchestration 143 00:07:23,080 --> 00:07:26,080 capability that's tailored to the customer journey runs in a 144 00:07:26,080 --> 00:07:28,880 serverless environment and then everything is protected by a 145 00:07:28,880 --> 00:07:30,840 sort of role based access control, right? 146 00:07:30,840 --> 00:07:34,160 So, so, so that's sort of the architecture differentiation in 147 00:07:34,160 --> 00:07:36,120 terms of capabilities. You know, as I mentioned, we're 148 00:07:36,120 --> 00:07:38,600 sitting here today, we're kind of the only solution that covers 149 00:07:38,800 --> 00:07:41,400 the full spectrum of customer identity needs right from a 150 00:07:41,400 --> 00:07:43,920 single product, single admin console. 151 00:07:44,160 --> 00:07:46,720 So if I put on my marketing hat, you know, I would like I say 152 00:07:46,720 --> 00:07:49,440 things like, hey, we can handle customer identity from sort of 153 00:07:49,440 --> 00:07:51,880 registration to the right to be forgotten, right? 154 00:07:51,880 --> 00:07:55,800 So the sort of beginning to end. And so you know, our, our 155 00:07:55,800 --> 00:07:58,880 capabilities cover things like registration, self-service, 156 00:07:59,240 --> 00:08:02,560 adaptive auth, identity verification, consent 157 00:08:02,560 --> 00:08:04,760 management, fraud detection, right. 158 00:08:05,160 --> 00:08:08,720 And then in terms of customer experience, you have to think 159 00:08:08,720 --> 00:08:11,760 about CIM systems as fundamentally workflow or 160 00:08:11,760 --> 00:08:14,920 journey systems, right? You, you have to build a whole 161 00:08:14,920 --> 00:08:19,600 layer on top of that IM tech around generating flexible flow 162 00:08:19,600 --> 00:08:21,720 based UIS for customers to interact with. 163 00:08:21,720 --> 00:08:24,640 So we think we've done a pretty good sort of differentiate job 164 00:08:24,640 --> 00:08:26,680 and sort of making it really easy to do this. 165 00:08:27,760 --> 00:08:30,240 So how do our customers measure success with the solution? 166 00:08:30,240 --> 00:08:32,320 It's a great question. I love, I love that question. 167 00:08:32,320 --> 00:08:36,360 There's, there's kind of two areas that I would focus on 168 00:08:36,360 --> 00:08:40,960 here. 1 is time to value, right? We, we talked about sort of 169 00:08:40,960 --> 00:08:45,200 deployments being long. This, you know, this is time to 170 00:08:45,200 --> 00:08:47,520 value is largely focused on things like deployment, sort of 171 00:08:47,520 --> 00:08:51,480 set up of the solution, things like customer journeys and sort 172 00:08:51,480 --> 00:08:53,800 of the migrations of of users, right. 173 00:08:54,840 --> 00:08:57,720 And the other is sort of ongoing KPIs around things like adoption 174 00:08:57,720 --> 00:09:00,320 metrics and customer engagement. And this is something where we 175 00:09:00,320 --> 00:09:05,840 really differentiate with our, with our dashboard, right? 176 00:09:05,840 --> 00:09:09,040 So, you know, you've got to make that really easy. 177 00:09:09,040 --> 00:09:12,280 You know, we, we can report on a cup on a few dozen different 178 00:09:12,280 --> 00:09:15,000 metrics, you know, across things like conversion rate, you know, 179 00:09:15,000 --> 00:09:17,800 how many visitors are actually converting into registered users 180 00:09:18,080 --> 00:09:21,680 on your platform, engagement rate, like how are your users? 181 00:09:22,040 --> 00:09:24,600 Are your user base sort of coming back every month, right? 182 00:09:25,680 --> 00:09:27,960 How long are they taking to register from the moment they 183 00:09:27,960 --> 00:09:30,640 click sign up to the moment they're redirected back to the 184 00:09:30,640 --> 00:09:32,160 customer portal? So these are things that you 185 00:09:32,160 --> 00:09:35,000 have to report on, right? So that so that they can your 186 00:09:35,000 --> 00:09:37,760 customers to sort of take those KP is back to their their 187 00:09:37,760 --> 00:09:40,520 management team and sort of show that they're, they're doing a 188 00:09:40,520 --> 00:09:42,920 good job, right. So let me ask you a real basic 189 00:09:42,920 --> 00:09:46,360 question because I feel like sometimes there's just double in 190 00:09:46,360 --> 00:09:49,200 the details. Is there a difference between 191 00:09:49,440 --> 00:09:53,640 consumer IAM and customer IAM or they really the the same thing? 192 00:09:54,280 --> 00:09:57,160 Yeah, there is a difference. I would say it's like a Venn 193 00:09:57,160 --> 00:10:00,440 diagram, right, where there's where there's overlap and 194 00:10:00,440 --> 00:10:03,600 there's and there's separate, there's separate capabilities, 195 00:10:03,600 --> 00:10:04,760 right. So a lot of the sort of 196 00:10:04,760 --> 00:10:08,600 fundamental IAM technology underneath, you know, things 197 00:10:08,600 --> 00:10:12,680 like federation, you know, open ID connect SAML, those sort of 198 00:10:12,680 --> 00:10:17,760 basic, you know, text are, are certain certainly within the 199 00:10:17,760 --> 00:10:20,480 middle of the Venn diagram between customer identity and 200 00:10:20,480 --> 00:10:23,920 and, and and workforce identity. But when you get into things 201 00:10:23,920 --> 00:10:27,560 like sort of customer metrics, you know, like how long is a 202 00:10:27,560 --> 00:10:30,840 user taking to register? Are you getting people that are 203 00:10:30,840 --> 00:10:35,440 sort of registering and walking away those there, there's not a 204 00:10:35,440 --> 00:10:36,760 whole lot of overlap there, right? 205 00:10:36,760 --> 00:10:39,920 Those are very customer specific metrics and those are things 206 00:10:39,920 --> 00:10:42,840 that, you know, you're sort of marketing team wants to know 207 00:10:43,240 --> 00:10:45,040 that your users are converting, right? 208 00:10:45,040 --> 00:10:47,400 That your users are making it in, in a, in a, in a workforce 209 00:10:47,400 --> 00:10:50,880 environment, you, you kind of have a, a captive audience, 210 00:10:50,880 --> 00:10:52,200 right? Like they're, you know, you're 211 00:10:52,200 --> 00:10:55,000 not going to have somebody that sort of refuses to register with 212 00:10:55,000 --> 00:10:58,560 the workforce, you know, access management system. 213 00:10:58,560 --> 00:11:01,320 And so there's a, you know, that there's that sort of where the, 214 00:11:01,440 --> 00:11:03,880 the sort of the, you know, on the outside of the, of the Venn 215 00:11:03,880 --> 00:11:06,360 diagram, you know what I mean? But yeah, it's a good question. 216 00:11:07,160 --> 00:11:09,000 Yeah, Steven. So I thought it was real 217 00:11:09,000 --> 00:11:12,120 interesting that our, our conversation started going down 218 00:11:12,120 --> 00:11:15,760 this kind of marketing track because, you know, big thing I 219 00:11:15,760 --> 00:11:19,920 wanted to talk about today was AI, of course, but even before 220 00:11:19,920 --> 00:11:22,960 that, it's just the topic of identity security. 221 00:11:23,360 --> 00:11:27,840 And you know, we had this episode scheduled for several 222 00:11:27,840 --> 00:11:31,080 months in advance. I've been in communication with 223 00:11:31,080 --> 00:11:35,240 Keith. And then just recently Gartner 224 00:11:35,520 --> 00:11:40,320 kind of identified you guys in the, in the paper, it's called 225 00:11:40,520 --> 00:11:43,760 emerging tech, tech landscape for startups and security 226 00:11:43,760 --> 00:11:46,920 software. And there is a digital identity 227 00:11:48,240 --> 00:11:52,000 column, if you will. And I think it's interesting 228 00:11:52,000 --> 00:11:56,080 because I'm I'm still not sure what is the difference between 229 00:11:56,080 --> 00:11:58,320 digital identity and identity security. 230 00:11:58,560 --> 00:12:00,760 I think that's maybe an episode for the future. 231 00:12:02,160 --> 00:12:07,360 But what do you think? It was about Trivacity that made 232 00:12:07,360 --> 00:12:10,640 Gardner sit up and take notice. Yeah. 233 00:12:10,640 --> 00:12:15,400 I mean, so Gardner's been talking about us, you know, for, 234 00:12:15,600 --> 00:12:19,160 for a while across a couple of different areas, which is kind 235 00:12:19,160 --> 00:12:22,640 of interesting. Journey time orchestration is 1 236 00:12:22,640 --> 00:12:25,800 where we've appeared in their research because that's a 237 00:12:25,800 --> 00:12:31,040 capability of our product, you know, sort of fraud and risk is 238 00:12:31,040 --> 00:12:33,360 one because we, you know, we have some fraud controls. 239 00:12:33,360 --> 00:12:36,360 We have things like identity verification and user 240 00:12:36,360 --> 00:12:38,440 experience, right, which is going to something that's going 241 00:12:38,440 --> 00:12:41,000 to make us stand out because it's, it hasn't been a focus of 242 00:12:41,000 --> 00:12:43,440 a lot of our, a lot of our competition. 243 00:12:44,560 --> 00:12:47,520 And also, you know, we're, we're, we're one of, you know, a 244 00:12:47,520 --> 00:12:50,800 small number of entrants into this space in the last five 245 00:12:50,800 --> 00:12:53,240 years that's taken significant investment, right? 246 00:12:53,240 --> 00:12:57,000 And, and especially 1 that's focused on the, on the customer 247 00:12:57,000 --> 00:13:01,400 identity problem and has made sort of, you know, customer 248 00:13:01,400 --> 00:13:04,640 traction, made some significant customer traction into, into 249 00:13:04,640 --> 00:13:07,560 this space. So I think it's great that 250 00:13:07,560 --> 00:13:09,280 they're taking notice. I mean, it's helping us with 251 00:13:09,280 --> 00:13:11,680 brand awareness. It's very crucial for our 252 00:13:11,680 --> 00:13:15,320 company, our stage to have that. And it's also kind of a sign 253 00:13:15,320 --> 00:13:18,040 that we're, we're probably getting things right, you know, 254 00:13:18,040 --> 00:13:22,160 from, from from the analyst point of view, you know, and in 255 00:13:22,160 --> 00:13:25,160 terms of sort of other, you know, other analysts activity. 256 00:13:26,160 --> 00:13:31,160 You know, we actually debuted as a leader in the Forster wave for 257 00:13:31,160 --> 00:13:35,640 CIM in 2022, which is just absolutely unheard of. 258 00:13:35,640 --> 00:13:39,720 Like for an, A, a start up to sort of debut as a, as a leader, 259 00:13:40,720 --> 00:13:44,320 you know, we were one of the only three leaders #2 after 260 00:13:44,320 --> 00:13:47,640 Fordrock at the time. There's another way of doing 261 00:13:47,640 --> 00:13:48,880 soon, I think. I believe. 262 00:13:48,880 --> 00:13:50,720 We believe it's going to be sometime in December. 263 00:13:51,800 --> 00:13:55,320 So you know, Gardner doesn't have a magic quadrant 264 00:13:55,320 --> 00:13:58,040 specifically for CIAM. But you know, like I mentioned, 265 00:13:58,040 --> 00:14:00,400 they're talking about us, you know, across other areas of 266 00:14:00,400 --> 00:14:03,000 research, which is great. And if I'm going to be, you 267 00:14:03,000 --> 00:14:06,600 know, if I'm going to be self-serving, you know, I'll say 268 00:14:06,600 --> 00:14:09,000 that, you know, such a young vendor debuting as a leader, you 269 00:14:09,000 --> 00:14:11,680 know, again on Forrester is just, is really a testament to 270 00:14:11,680 --> 00:14:14,880 the, to the strategy and just how good our product is. 271 00:14:15,360 --> 00:14:18,400 You know, I think you know our architecture, our breadth of 272 00:14:18,400 --> 00:14:21,560 capabilities, our focus on customers experience is really 273 00:14:21,560 --> 00:14:24,760 what's behind the interest from from the analysts. 274 00:14:25,520 --> 00:14:26,960 Yeah. So you talked about the 275 00:14:26,960 --> 00:14:30,520 isolation by design. I think that's just really 276 00:14:30,520 --> 00:14:34,120 interesting, but talk to us about why that's important from 277 00:14:34,120 --> 00:14:40,040 a identity security standpoint. Yeah, so I, I came from an 278 00:14:40,040 --> 00:14:43,560 incident response background. So I, I worked in network and 279 00:14:43,560 --> 00:14:47,680 endpoint security for a company called Mandiant, probably heard 280 00:14:47,680 --> 00:14:52,800 of, you know, later acquired by Google and you know, I sort of 281 00:14:52,800 --> 00:14:57,280 saw the the, the worst of the worst of, of breaches, right. 282 00:14:57,280 --> 00:15:02,880 So when you know you're asking a, a large org, you know a large 283 00:15:02,880 --> 00:15:06,200 financial institution, a large healthcare business to sort of 284 00:15:06,200 --> 00:15:10,560 move their their data to the cloud, you better be protecting 285 00:15:10,560 --> 00:15:13,480 it, right? So one of our sort of 286 00:15:13,480 --> 00:15:15,720 fundamental design principles was we're going to, we're going 287 00:15:15,720 --> 00:15:18,360 to give an isolated instance of our product to each of our 288 00:15:18,360 --> 00:15:20,440 customers, right? That means that means that if 289 00:15:20,440 --> 00:15:23,440 one of our customers is breached, the blast radius 290 00:15:23,440 --> 00:15:27,240 between one customer to the next is, is null, right? 291 00:15:27,240 --> 00:15:30,720 Like there's no, there's no shared infrastructure, no shared 292 00:15:30,720 --> 00:15:34,920 keys, databases, services between any two of our 293 00:15:34,920 --> 00:15:37,400 customers, right? And and that's, that's 294 00:15:38,040 --> 00:15:41,160 absolutely critical in customer data, customer identity, because 295 00:15:41,160 --> 00:15:43,760 you're storing, you know, sensitive customer data, you're 296 00:15:43,760 --> 00:15:47,080 storing passwords, you're storing MFA, you know, MFA 297 00:15:47,080 --> 00:15:50,560 credentials like this is really, really sensitive data that 298 00:15:50,560 --> 00:15:52,120 you're storing. So you have to sort of take a 299 00:15:52,120 --> 00:15:57,520 really strong security, you know, posture with how you 300 00:15:57,520 --> 00:16:00,360 design these systems. There was a LinkedIn post that 301 00:16:00,360 --> 00:16:04,360 Jim actually made about why I am tech companies aren't really 302 00:16:04,360 --> 00:16:06,720 getting into the generative AI space of their products yet. 303 00:16:06,720 --> 00:16:08,720 I think it's going to be a little bit dated, Jim, no 304 00:16:08,720 --> 00:16:10,880 offense, because I think we're starting to see it more out 305 00:16:10,880 --> 00:16:12,360 there. But what? 306 00:16:12,600 --> 00:16:16,000 And you responded back to that, like what prompted that response 307 00:16:16,000 --> 00:16:18,000 back? Was it just familiar with Jim or 308 00:16:18,000 --> 00:16:19,800 you're like, oh, hold on a second buddy, you're way off 309 00:16:19,800 --> 00:16:23,000 base here. Well, I mean one, I, I do like 310 00:16:23,000 --> 00:16:25,520 you guys. So I, you know, I, I do. 311 00:16:25,520 --> 00:16:26,960 We'll get you everywhere on the show. 312 00:16:27,920 --> 00:16:30,120 Yeah. So, so that was certainly one, 313 00:16:30,800 --> 00:16:33,160 but I don't know it just it, it was interesting. 314 00:16:33,160 --> 00:16:35,960 It was, it was during a time when we were sort of deeply 315 00:16:35,960 --> 00:16:38,400 pondering that very question, right. 316 00:16:38,760 --> 00:16:42,560 So I sort of struck me as a very relevant question from from Jim, 317 00:16:42,720 --> 00:16:46,280 kind of considering where we are in the space, where we are in 318 00:16:46,280 --> 00:16:49,120 sort of the development of AI, you know, where we are in the 319 00:16:49,120 --> 00:16:52,640 identity space. You know, every sort of product 320 00:16:52,640 --> 00:16:56,560 leader right now is having to make a decision as to whether 321 00:16:56,560 --> 00:16:59,240 integrating, you know, generative AI and into their 322 00:16:59,240 --> 00:17:03,160 platform makes sense, right? And, and, and IAM products are 323 00:17:03,160 --> 00:17:05,560 no, are no different, you know, than that. 324 00:17:05,560 --> 00:17:10,319 So that was kind of the reason. Jim, Jim, you wrote this article 325 00:17:10,319 --> 00:17:11,640 and I don't want to leave people behind. 326 00:17:11,640 --> 00:17:14,040 So you kind of give like a gist of it so you can catch people 327 00:17:14,040 --> 00:17:17,119 who are listening or watching. Yeah, it wasn't really so much 328 00:17:17,119 --> 00:17:22,680 an article as much as it was, you know, a a true question. 329 00:17:22,680 --> 00:17:27,359 Like it seems like every product is now baking in AI. 330 00:17:27,640 --> 00:17:31,480 But in the IM space, at least, what I'm seeing is, you know, 331 00:17:32,280 --> 00:17:38,080 versions of AI are more big data analytics, machine learning. 332 00:17:38,400 --> 00:17:41,960 They're not generative AI like what you get when you go out to 333 00:17:42,200 --> 00:17:47,200 open AI or, you know, Gemini or something like that. 334 00:17:47,560 --> 00:17:51,360 And, you know, I throw this one back to Steven because, you 335 00:17:51,360 --> 00:17:52,720 know, is that what you see as well? 336 00:17:52,720 --> 00:17:56,080 Like there's a difference between machine learning and big 337 00:17:56,080 --> 00:18:00,880 data analytics and AI, right? Yeah, I think, yeah, definitely 338 00:18:00,920 --> 00:18:03,880 you need to stick, take a step back and and and sort of think 339 00:18:03,880 --> 00:18:06,600 about like what is the intersection between AI and ML, 340 00:18:06,600 --> 00:18:08,280 right. Like the media often gets this 341 00:18:08,280 --> 00:18:12,640 wrong, as you, as you've likely noticed, you know, and I think I 342 00:18:12,640 --> 00:18:15,480 think the way to position is that, you know, AI is just sort 343 00:18:15,480 --> 00:18:20,880 of overarching term of which machine learning is a subset, 344 00:18:20,960 --> 00:18:22,640 right? It's, it's correct to say that 345 00:18:22,640 --> 00:18:26,520 ML is a form of AI, but ML has been sort of historically 346 00:18:26,520 --> 00:18:29,760 focused on finding patterns and data and getting smarter over 347 00:18:29,760 --> 00:18:32,760 time so that you can make decisions on what it finds and 348 00:18:32,760 --> 00:18:36,200 protects. So if you think about ML, we've 349 00:18:36,200 --> 00:18:38,880 actually like, we specifically Stervastity has had it in our 350 00:18:38,880 --> 00:18:40,480 product for a number of years, right? 351 00:18:40,480 --> 00:18:43,080 Specifically around the behavior analytics problem. 352 00:18:44,120 --> 00:18:46,800 So let's sort of watching behavior, user behavior 353 00:18:46,800 --> 00:18:48,920 detecting when there's an anomaly, you know, sort of 354 00:18:48,920 --> 00:18:50,880 making access decisions based on that. 355 00:18:51,040 --> 00:18:55,240 And to be honest, like this ML based behavior analytics is, is 356 00:18:55,240 --> 00:18:58,640 by no means a new technology. The first, the first UEUEBA 357 00:18:58,640 --> 00:19:02,200 vendors were got their start in what 2014? 358 00:19:02,200 --> 00:19:06,640 And I mean, that's an eon in, in technology terms. 359 00:19:07,360 --> 00:19:09,480 So it's, but it's also means it's very tried and true 360 00:19:09,480 --> 00:19:12,640 technology for, for, for what it's designed for right now. 361 00:19:12,640 --> 00:19:15,240 If you're talking specifically about generative AI or Gen. 362 00:19:15,240 --> 00:19:19,680 AI, that's what we sort of most recently added to our platform. 363 00:19:19,680 --> 00:19:23,320 And what I think a lot of you know, product practitioners are 364 00:19:23,320 --> 00:19:27,240 sort of pondering right now as to what makes sense for them. 365 00:19:28,600 --> 00:19:31,800 You know, if you talk about Gen. AI as soaring sort of also being 366 00:19:31,800 --> 00:19:36,120 a subset of the overall AI space, it also makes use of 367 00:19:36,120 --> 00:19:37,400 machine learning. Gen. 368 00:19:37,400 --> 00:19:39,560 Gen. AI models are designed to emit 369 00:19:39,920 --> 00:19:45,080 or produce data versus simply analyze and, and classify data. 370 00:19:45,600 --> 00:19:48,200 But they also have other, other elements that sort of fall into 371 00:19:48,200 --> 00:19:50,640 that bucket. You know, a good example would 372 00:19:50,640 --> 00:19:52,640 be like natural language processing, right? 373 00:19:52,640 --> 00:19:57,320 The chatbot aspect of of Gen. AI systems is a is a somewhat 374 00:19:57,320 --> 00:19:59,520 separate space and it's, you know, it's all about determining 375 00:19:59,520 --> 00:20:04,440 user intent. Or what they call grounding in 376 00:20:04,440 --> 00:20:06,120 the in the Gen. AI space. 377 00:20:06,120 --> 00:20:10,280 So you know, when you sort of couple a prompt that you're 378 00:20:10,280 --> 00:20:13,040 sending to an LLM with live data, right? 379 00:20:13,040 --> 00:20:18,080 You, you, you might have heard this referred to as retrieval 380 00:20:18,080 --> 00:20:21,920 augmented generation or RAG, which is the way that a lot of 381 00:20:21,920 --> 00:20:25,040 products right now are implementing AI into their 382 00:20:25,040 --> 00:20:26,840 system. So, you know, we circle back, 383 00:20:26,840 --> 00:20:29,080 like I say, we've had ML in our product for some time. 384 00:20:29,280 --> 00:20:31,080 Gen. AI was a was a recent addition. 385 00:20:31,080 --> 00:20:34,520 I hope that was a pretty clear description of of how I see see 386 00:20:34,520 --> 00:20:36,360 them differently. No, it's great. 387 00:20:36,360 --> 00:20:38,440 I mean, I feel like we can, we can really dive into this 388 00:20:38,440 --> 00:20:40,200 because we kind of joke all the time that this we should turn 389 00:20:40,200 --> 00:20:43,560 this into like AI at the center of the way the this is moving. 390 00:20:44,560 --> 00:20:50,720 So my definition of AI changed when I first saw the open AILLM 391 00:20:50,720 --> 00:20:55,400 sort of released and I was like, oh wow, OK, That AI before that 392 00:20:55,400 --> 00:20:58,080 really was like machine learning and sort of like behind the 393 00:20:58,080 --> 00:20:59,600 curtains. And now all of a sudden we've 394 00:20:59,600 --> 00:21:03,000 got a text based interface at the time and now it's even voice 395 00:21:03,480 --> 00:21:06,040 that you can, you know, buddy up with and ask questions. 396 00:21:06,040 --> 00:21:08,080 Now, do you trust it or not? It's different. 397 00:21:08,120 --> 00:21:10,360 It's a different, you know, answer or a question. 398 00:21:10,360 --> 00:21:14,760 But did your definition or your thought process around AI change 399 00:21:14,760 --> 00:21:16,880 when you saw that or you'd already seen that coming? 400 00:21:16,880 --> 00:21:20,320 I mean. Oh, I mean 100%. 401 00:21:20,320 --> 00:21:24,160 I mean, the, I was, you know, I was as I was sort of, you know, 402 00:21:24,160 --> 00:21:26,400 thinking about, you know, putting some notes, some 403 00:21:26,400 --> 00:21:28,840 thoughts and notes together today for, for the discussion 404 00:21:28,840 --> 00:21:31,800 today. I, I realized that chat TBT was 405 00:21:31,800 --> 00:21:34,400 only real was only released like in 2022. 406 00:21:34,400 --> 00:21:37,720 It was like 2 years ago. I mean, that's crazy that it 407 00:21:37,720 --> 00:21:40,480 seems like, it seems like it's been like 20 years, you know 408 00:21:40,480 --> 00:21:43,440 what I mean? And so, yeah, I mean, I think I, 409 00:21:43,520 --> 00:21:48,360 I think I was just as shocked as as a lot of the folks in, in, in 410 00:21:48,360 --> 00:21:51,920 the technology space where the sort of power of these of these 411 00:21:51,920 --> 00:21:57,440 systems of 100%, it completely shook my, my, my understanding 412 00:21:57,440 --> 00:22:01,280 of the space 100%. So let's get more into that AI 413 00:22:01,280 --> 00:22:04,120 because you know, I mentioned the website Stravacity dot AI. 414 00:22:04,120 --> 00:22:08,200 It's literally in the name. What are what are some of the 415 00:22:08,200 --> 00:22:11,520 things that we can expect if I if I pull up trivacity? 416 00:22:11,520 --> 00:22:13,240 Give me some examples of like use cases. 417 00:22:14,240 --> 00:22:17,760 Yeah. So you know, on the benefits, 418 00:22:17,760 --> 00:22:20,200 you know sort of like the benefits of pulling this into 419 00:22:20,200 --> 00:22:24,520 our product, there's a couple things that really just line up 420 00:22:24,520 --> 00:22:27,640 with how we have built the product from the beginning, 421 00:22:27,680 --> 00:22:30,320 right. What one of one of those is sort 422 00:22:30,320 --> 00:22:34,760 of lowering the bar, you know, of entry to, to IAM systems, 423 00:22:34,760 --> 00:22:37,400 right? We, we always say that you kind 424 00:22:37,400 --> 00:22:40,240 of, you don't need to, you shouldn't have to have a PhD in 425 00:22:40,240 --> 00:22:42,880 IAM to set up a customer journey, right? 426 00:22:43,800 --> 00:22:48,240 And so, you know, Gen. AI makes a really good companion 427 00:22:48,240 --> 00:22:51,560 to work with as you're sort of configuring and maintaining 428 00:22:51,560 --> 00:22:55,360 your, your IAM systems. You know, things like context 429 00:22:55,360 --> 00:22:59,800 sensitive help where you're asking questions of an assistant 430 00:22:59,800 --> 00:23:02,280 on how to, you know, that AI assist on how to do certain 431 00:23:02,280 --> 00:23:06,280 tasks within the platform. Like, you know, we can tell 432 00:23:06,280 --> 00:23:08,080 where you are in the admin console. 433 00:23:08,160 --> 00:23:12,680 We can send along pertinent contacts and our docs to the LLM 434 00:23:12,920 --> 00:23:16,400 to craft a very specific answer. Like, so hey, how do I set up 435 00:23:16,400 --> 00:23:20,360 bot detection? Or how do I change my password 436 00:23:20,360 --> 00:23:23,720 policy to be more restrictive? You know, like we can really 437 00:23:23,720 --> 00:23:25,240 help with context sensitive help, right? 438 00:23:25,240 --> 00:23:28,160 That's a, that's a huge, there's one thing we wanted to build for 439 00:23:28,160 --> 00:23:31,800 a while is constant context sensitive help. 440 00:23:31,800 --> 00:23:36,760 And Eugenia is a great, a great way, you know, to to sort of 441 00:23:36,760 --> 00:23:42,160 facilitate that interpretation of data is another one that we 442 00:23:42,160 --> 00:23:45,240 thought was really important. You know, like I am isn't just 443 00:23:45,240 --> 00:23:47,960 about like MFA and federation, right? 444 00:23:48,240 --> 00:23:50,120 We can also generate a lot of data. 445 00:23:51,240 --> 00:23:53,480 It's often hard for less technical people to crock, 446 00:23:53,640 --> 00:23:57,120 right? And you know, specifically in in 447 00:23:57,120 --> 00:24:01,160 customer identity, you know, where some of your users of the 448 00:24:01,160 --> 00:24:04,280 platform may actually be in fact customer service 449 00:24:04,280 --> 00:24:09,320 representatives, CSR's who are deeply technical in IAM areas, 450 00:24:09,680 --> 00:24:11,360 but they might actually be actively on the phone with a 451 00:24:11,360 --> 00:24:13,120 customer and they're trying to cut, you know, troubleshoot a 452 00:24:13,120 --> 00:24:17,840 customer issue. So roll ups of these types of 453 00:24:17,840 --> 00:24:20,440 these log data of this customer journey data can actually be 454 00:24:20,440 --> 00:24:22,880 really helpful in getting to an answer quickly when a customer's 455 00:24:22,880 --> 00:24:25,000 unhappy, right? We can say like, oh, you're this 456 00:24:25,000 --> 00:24:27,080 user is struggling to get through the identity 457 00:24:27,080 --> 00:24:30,320 verification process or they're not able to finish their 458 00:24:30,320 --> 00:24:32,880 password enrollment, you know, those kind of things, right? 459 00:24:33,120 --> 00:24:36,040 So that's really helping there. And then also like one of the 460 00:24:36,040 --> 00:24:38,440 things we wanted to really focus on and I think is really big 461 00:24:38,440 --> 00:24:42,320 benefit is, is sort of the guidance towards best practice 462 00:24:42,320 --> 00:24:46,480 or optimization aspects. So, you know, there's a lot of 463 00:24:46,480 --> 00:24:49,040 best practice out there, password guidelines, adaptive 464 00:24:49,040 --> 00:24:52,960 access policies, you know, how you know, what if you could 465 00:24:52,960 --> 00:24:56,880 guide A-Team or a user towards improving their posture based on 466 00:24:56,880 --> 00:25:00,280 what they're actually seeing in the product, Specifically on the 467 00:25:00,280 --> 00:25:03,400 CIAM side, What if you could got a, you know, got a team towards 468 00:25:03,400 --> 00:25:06,120 optimizing customer experience or improving, you know, 469 00:25:06,720 --> 00:25:09,000 conversion rates, spice, what you're seeing in the logs, what 470 00:25:09,000 --> 00:25:11,200 you're seeing in the data, what you're seeing in the, in the 471 00:25:11,200 --> 00:25:14,000 dashboard. So those are, you know, those 472 00:25:14,000 --> 00:25:16,840 are kind of those are kind of the focuses and kind of the 473 00:25:16,840 --> 00:25:20,640 things that you can kind of do with with our, our, our current 474 00:25:20,640 --> 00:25:26,320 AI assistant in the product. Yeah, I'm kind of I'm interested 475 00:25:26,320 --> 00:25:29,720 in in, you know, visualizing in my head. 476 00:25:29,960 --> 00:25:32,320 How does that take place? So that last part that you're 477 00:25:32,320 --> 00:25:37,160 talking about where it's like, all right, I want the system can 478 00:25:37,520 --> 00:25:44,280 make suggestions to your client in terms of access policies or 479 00:25:44,280 --> 00:25:46,120 things. How does that manifest? 480 00:25:47,160 --> 00:25:53,000 So when the user comes along and they're looking at a particular 481 00:25:53,280 --> 00:25:58,320 portion of the admin console and they ask a question, we do a 482 00:25:59,120 --> 00:26:02,360 form of retrieval augmented generation, which is RAG, which 483 00:26:02,360 --> 00:26:06,280 is just a, a, a really complex way to say that we take the 484 00:26:06,280 --> 00:26:10,840 prompt, we take the context of what they're looking at in the, 485 00:26:10,880 --> 00:26:15,040 in the admin console and we take a relevant section of our docs. 486 00:26:15,320 --> 00:26:18,040 And we send all of that to the LLM at the same time. 487 00:26:18,040 --> 00:26:22,920 And we say, give me an answer built from this prompt, this 488 00:26:22,920 --> 00:26:28,000 context, and this documentation. Now this is obviously very, very 489 00:26:28,080 --> 00:26:33,120 simplified version of a complex thing, but that's what happens. 490 00:26:33,120 --> 00:26:37,040 And then the LLM has what it's what it needs to answer the 491 00:26:37,040 --> 00:26:40,320 question in a way that's very specific to our product and 492 00:26:41,200 --> 00:26:45,080 relevant to the context of what the user is, is, is staring at 493 00:26:45,080 --> 00:26:47,040 at that at that point in time, right. 494 00:26:47,400 --> 00:26:51,040 And this is a very, this is a pattern that you will see in the 495 00:26:51,040 --> 00:26:52,800 Gen. AI space right now of people 496 00:26:52,800 --> 00:26:56,640 doing these, these rag based systems that are marrying a 497 00:26:56,640 --> 00:27:00,880 prompt with context, with documentation, with maybe stuff 498 00:27:00,880 --> 00:27:03,560 on the web, right? Open AI can do that when you 499 00:27:03,640 --> 00:27:06,680 when you ask a question about current events, it's going to go 500 00:27:06,680 --> 00:27:10,200 out and search the web, search news articles for relevant stuff 501 00:27:10,200 --> 00:27:13,640 and then give you a context relevant answer. 502 00:27:13,640 --> 00:27:16,560 That is another form of rag, right, that that you're seeing 503 00:27:16,840 --> 00:27:20,080 Open AI implement in their ChatGPT tool. 504 00:27:20,080 --> 00:27:24,560 So it's a common pattern. That seems like a a really 505 00:27:24,720 --> 00:27:29,600 clever way to help with the security of the model itself and 506 00:27:29,720 --> 00:27:32,320 reduce the chance of a hallucination, right? 507 00:27:32,320 --> 00:27:34,840 I think everyone's kind of concerned about will the AI give 508 00:27:34,840 --> 00:27:36,600 me a correct answer and can I trust it? 509 00:27:36,600 --> 00:27:38,200 And what if it comes up with something to put you off the 510 00:27:38,200 --> 00:27:39,920 wall? But it sounds like if you're 511 00:27:39,920 --> 00:27:43,400 able to limit the input into that model and have it just be 512 00:27:43,400 --> 00:27:46,920 based on your documentation, you know the chances of it going off 513 00:27:46,920 --> 00:27:49,880 the rails are probably a little bit, you know better for it not 514 00:27:49,880 --> 00:27:54,120 happening. 100% and like, and that, and that goes back to the, 515 00:27:54,240 --> 00:27:58,240 the, the comment I made about grounding, right, grounding is, 516 00:27:58,240 --> 00:28:01,400 is, is a, is a necessary technology and sort of 517 00:28:01,400 --> 00:28:03,480 implementing Gen. AI systems, Gen. 518 00:28:03,560 --> 00:28:06,760 AI products, because basically what you're doing is you're, 519 00:28:07,280 --> 00:28:09,840 you're, you're closing those guard rails in a little bit, 520 00:28:09,840 --> 00:28:12,840 right And you're giving, you're giving an LLM more information 521 00:28:12,840 --> 00:28:16,280 to give you an accurate answer, right. 522 00:28:16,480 --> 00:28:19,240 And if you think about like, I love the hallucination topic. 523 00:28:19,520 --> 00:28:21,760 I, I think it's AI think it's, I think it's great that you 524 00:28:21,760 --> 00:28:23,440 brought that up. I mean there's, there's been 525 00:28:23,440 --> 00:28:26,560 some really fun stories about that in the, in the media. 526 00:28:27,520 --> 00:28:31,200 And I think, I think it's really useful for, for, you know, your, 527 00:28:31,200 --> 00:28:35,640 your listeners to understand like what a hallucination is 528 00:28:35,640 --> 00:28:39,880 and, and why you, you need to understand why they happen. 529 00:28:39,880 --> 00:28:42,560 I, and I think if you, if you look at what an LLM 530 00:28:42,560 --> 00:28:45,840 fundamentally is, it's, it's, it's a probability machine, 531 00:28:45,880 --> 00:28:47,840 right? I mean, I'll probably get 532 00:28:47,840 --> 00:28:51,400 blowback from your users for that simplification, but maybe 533 00:28:51,400 --> 00:28:53,640 it's better to say that they're sort of a, they're sort of 534 00:28:53,640 --> 00:28:55,520 fundamentally probabilistic in nature. 535 00:28:56,600 --> 00:28:58,640 So what they're trying to do is they're trying to figure out 536 00:28:58,640 --> 00:29:03,120 what the next word, character phrase is based on the context 537 00:29:03,120 --> 00:29:05,440 that they know about what you, what you've sent it, right? 538 00:29:05,440 --> 00:29:10,000 So, you know, a hallucination can kind of be thought of as a, 539 00:29:10,000 --> 00:29:12,520 as a statement that's statistically probable, but 540 00:29:12,520 --> 00:29:14,920 factually incorrect. I, I, I like that. 541 00:29:15,160 --> 00:29:17,120 I read that statement somewhere. I was like, that's perfect way 542 00:29:17,120 --> 00:29:18,840 to describe it. Statistically probable, 543 00:29:18,960 --> 00:29:21,680 factually incorrect, right. And again, this is one of the 544 00:29:21,680 --> 00:29:25,240 reasons that people are building grounding systems around AIS 545 00:29:25,240 --> 00:29:28,720 like like RAG, because it's a safety mechanism, It's a 546 00:29:28,720 --> 00:29:30,720 security mechanism, like you said, right? 547 00:29:30,720 --> 00:29:33,640 To make it more, make the answer more reliable, make it more 548 00:29:33,640 --> 00:29:36,760 accurate. It's certainly a concern for us 549 00:29:36,760 --> 00:29:39,480 as we were, as we were building the product, building it into 550 00:29:39,480 --> 00:29:41,120 our product. And it's also why you see 551 00:29:41,120 --> 00:29:44,040 disclaimers like on chatbot systems. 552 00:29:44,040 --> 00:29:46,000 It's like, hey, LLMS can make mistakes. 553 00:29:46,000 --> 00:29:49,120 Please, please verify your answers because hallucinations 554 00:29:49,120 --> 00:29:51,440 are and will continue to be a problem even with grounding 555 00:29:51,440 --> 00:29:55,960 systems around around, you know, the, the implementation of 556 00:29:55,960 --> 00:29:58,400 these. So yeah, I think it's, I think 557 00:29:58,400 --> 00:30:02,920 it's a great question. Is are hallucinations the the 558 00:30:02,920 --> 00:30:06,960 biggest threat to the success of an AI right now and trying to 559 00:30:06,960 --> 00:30:09,600 put guard rails around it so that you're not giving incorrect 560 00:30:09,600 --> 00:30:10,920 information? Or is there something else that 561 00:30:10,920 --> 00:30:13,000 keeps you up at night? I mean, there's, you know, 562 00:30:13,000 --> 00:30:15,720 there's a hallucinations are definitely a challenge. 563 00:30:15,720 --> 00:30:19,160 I don't, I don't, I think you can kind of especially when 564 00:30:19,160 --> 00:30:21,840 you're not, you're not sort of implementing what you know, what 565 00:30:21,840 --> 00:30:25,760 I would consider a public API, a public AI system like like 566 00:30:25,760 --> 00:30:28,760 ChatGPT. Ours is very specific to our 567 00:30:28,760 --> 00:30:31,280 product. So we can really control what 568 00:30:31,280 --> 00:30:34,760 sort of goes into it. But you have all kinds of other 569 00:30:34,760 --> 00:30:38,280 challenges like, you know, prompt injection is one, right? 570 00:30:38,280 --> 00:30:40,240 Because if you know, it's not all that different than like a 571 00:30:40,240 --> 00:30:44,640 sequel injection where you're, you're, you're taking a prompt, 572 00:30:44,640 --> 00:30:47,040 you're formulating it into a larger prompt and you're 573 00:30:47,040 --> 00:30:49,800 injecting that in the LLM. And a smart user might figure 574 00:30:49,800 --> 00:30:53,240 out what the format of your prompt is and sort of break out 575 00:30:53,240 --> 00:30:55,800 of it in some way, right? And try to get data out of your 576 00:30:55,800 --> 00:30:58,680 system. But you can also put, you know, 577 00:30:58,920 --> 00:31:00,960 a system like ours. You can also put guard rails 578 00:31:00,960 --> 00:31:04,160 further down the road so that you're not able to return or 579 00:31:04,160 --> 00:31:06,640 you're only talking to certain APIs that are read only. 580 00:31:06,640 --> 00:31:08,360 And you're, you're, you know, you're, you're putting some 581 00:31:08,360 --> 00:31:13,920 security guidelines past the LLM point within the product. 582 00:31:13,920 --> 00:31:16,120 So those kind of things keep me up at night. 583 00:31:16,120 --> 00:31:18,400 I honestly, I think the other thing that keeps me up at night 584 00:31:18,400 --> 00:31:21,800 is sort of the, the, the cost around these systems, right? 585 00:31:22,640 --> 00:31:26,360 That's, that was a, the first thing, the first one of the sort 586 00:31:26,360 --> 00:31:28,800 of the first things that I thought about when, when we said 587 00:31:28,800 --> 00:31:30,880 let's, let's put this into our product was like, Oh my gosh, 588 00:31:30,880 --> 00:31:32,240 how much is this going to cost? Right? 589 00:31:32,240 --> 00:31:36,960 Like it's compute and memory. It's you requires you to sort of 590 00:31:36,960 --> 00:31:40,840 have specialized GPU heavy instances in the in the cloud. 591 00:31:41,240 --> 00:31:43,880 And so you know that when you're sort of talking about large 592 00:31:43,880 --> 00:31:47,880 scale systems where a user, you know, where you know, you have 593 00:31:47,880 --> 00:31:51,880 customers that have, you know, 10s of millions of users, you 594 00:31:51,880 --> 00:31:53,760 know, a large portion of which login every month. 595 00:31:53,760 --> 00:31:56,400 Like are you are you being careful about how you're 596 00:31:56,400 --> 00:32:00,160 leveraging, leveraging the tech right with regards to large 597 00:32:00,160 --> 00:32:03,560 scale deployments? You know, that's that was that 598 00:32:03,560 --> 00:32:07,400 was sort of a big worry for me, less of a worry now, but but was 599 00:32:07,400 --> 00:32:10,040 a big worry in the at the at the, you know, the the start. 600 00:32:10,320 --> 00:32:14,240 Yeah, you're starting to get into where I think the the big 601 00:32:14,240 --> 00:32:18,880 value and the big threat are. So I think the big value is 602 00:32:18,880 --> 00:32:25,520 being able to, you know, enable say marketing you brought up in 603 00:32:25,520 --> 00:32:29,400 the marketing example for business users to query the data 604 00:32:29,720 --> 00:32:33,040 so they can start saying, OK, well, how many people logged in 605 00:32:33,320 --> 00:32:36,400 in the last 24 hours? That's one level. 606 00:32:36,480 --> 00:32:40,600 Another level might be now let's branch out into information 607 00:32:40,600 --> 00:32:42,560 that's on the Internet. How many people logged in in the 608 00:32:42,560 --> 00:32:47,480 last 24 hours from somewhere that is was raining, you know, 609 00:32:47,760 --> 00:32:52,840 and you know, so the, the concern it brings up is like, 610 00:32:52,840 --> 00:32:57,200 OK, how far can it extend? Could I, you know, what do I 611 00:32:57,200 --> 00:33:01,240 have to do to create security boundaries to make sure that I 612 00:33:01,240 --> 00:33:05,560 don't say, all right, what is my competitor doing on this system? 613 00:33:05,560 --> 00:33:09,560 How many users do they have? And maybe I do that in some kind 614 00:33:09,560 --> 00:33:14,760 of like less obvious way, right? Maybe I know that my competitor 615 00:33:14,760 --> 00:33:18,320 is using this system. So I, I start asking questions 616 00:33:18,320 --> 00:33:22,120 like more at like an industry level, like what do 617 00:33:22,800 --> 00:33:26,280 organizations that are in my industry, how many users do they 618 00:33:26,280 --> 00:33:29,200 normally have? Things like how many, how much, 619 00:33:29,200 --> 00:33:32,280 how many times do people log in the last 24 hours for other 620 00:33:32,280 --> 00:33:34,120 organizations that are in my industry? 621 00:33:34,320 --> 00:33:39,800 So, Mike, I think that's the real power would be to be able 622 00:33:39,800 --> 00:33:44,520 to ask, you know, reporting style questions that it could 623 00:33:44,720 --> 00:33:48,240 really think about and maybe you could even pitch a scenarios 624 00:33:48,240 --> 00:33:53,840 like, all right, well, if a hurricane took out power over 625 00:33:54,160 --> 00:33:58,080 the, you know, this large portion, Jeff and I certainly 626 00:33:58,080 --> 00:34:01,680 know that problem. How would that, how would that 627 00:34:02,120 --> 00:34:07,040 impact my logins? I, I knew this as I am, this is, 628 00:34:07,280 --> 00:34:09,440 you know, I can't really basically go as far as to say 629 00:34:09,600 --> 00:34:13,360 how would that impact my sales? But maybe, you know, something 630 00:34:13,360 --> 00:34:17,679 like that. And then I think the, the big 631 00:34:17,679 --> 00:34:22,159 threat is the more powerful that becomes, the more security 632 00:34:22,159 --> 00:34:26,480 becomes an issue, which probably goes to like, how did you guys 633 00:34:26,480 --> 00:34:29,840 go about deploying this? Is this are you leveraging like 634 00:34:29,840 --> 00:34:32,760 the APIs of some big tech platform? 635 00:34:33,600 --> 00:34:37,960 Did you develop your own AI or like how did you build AI into 636 00:34:37,960 --> 00:34:39,840 your product? Yeah. 637 00:34:39,840 --> 00:34:47,239 So again, we, you know, we we focused on isolation from the 638 00:34:47,239 --> 00:34:50,560 get go, right, which actually put us in a pretty good position 639 00:34:50,760 --> 00:34:54,400 for implementing this technology into the platform. 640 00:34:54,400 --> 00:34:59,480 So we are, we're leveraging an existing LLM, but it's local to 641 00:34:59,480 --> 00:35:02,600 our customer environment, our individual customer environment. 642 00:35:02,600 --> 00:35:06,960 So they get their own instance of our LLM within, within the 643 00:35:06,960 --> 00:35:10,080 customer within their own, you know, product instance. 644 00:35:10,400 --> 00:35:13,440 So we're not using a public LLM like open AI, right? 645 00:35:13,720 --> 00:35:16,680 And as soon as you open that door where, you know, these 646 00:35:16,680 --> 00:35:20,480 systems are able to sort of go out to the Internet or, you 647 00:35:20,480 --> 00:35:24,360 know, go out to a third party system, you're, you know, that 648 00:35:24,360 --> 00:35:28,120 opens a lot more, a lot more risk than having a system that's 649 00:35:28,120 --> 00:35:31,160 sort of self-contained within, within our admin console within 650 00:35:31,160 --> 00:35:33,760 our own customer instance. It only has access to that 651 00:35:33,760 --> 00:35:36,640 customer data within within that instance. 652 00:35:36,640 --> 00:35:41,720 So from in terms of like a, an isolation or sort of a, you 653 00:35:41,720 --> 00:35:45,360 know, securing a customer perspective, we, we were kind of 654 00:35:45,360 --> 00:35:48,120 already in a good position for that right in the way we wanted 655 00:35:48,120 --> 00:35:52,000 it to implement that. And there's, there's a lot of 656 00:35:52,920 --> 00:35:57,160 local open source models out there that are extremely 657 00:35:57,160 --> 00:36:01,480 powerful and well suited to RAG systems. 658 00:36:01,480 --> 00:36:05,600 Like what we like, what we built with Instrabassity, Everything 659 00:36:05,600 --> 00:36:08,400 else, you know, around it we built is custom, right? 660 00:36:08,400 --> 00:36:12,560 The actual RAG implementation, you know, the user interface, 661 00:36:12,560 --> 00:36:16,160 you know, the sort of, you know, AI assist like interface we have 662 00:36:16,160 --> 00:36:17,480 in our admin console, things like that. 663 00:36:17,480 --> 00:36:21,520 That's all, that's all custom. But yeah. 664 00:36:22,120 --> 00:36:26,160 And then to to kind of like agree with the point that I was 665 00:36:26,160 --> 00:36:29,800 making where like definitely I think the things that you talked 666 00:36:29,800 --> 00:36:34,240 about a lot of power in that and a lot of benefit in having 667 00:36:34,560 --> 00:36:39,760 contextual base help. But I think kind of that next 668 00:36:39,760 --> 00:36:44,600 frontier is, you know, just leaving it open to non-technical 669 00:36:44,600 --> 00:36:49,440 people to be able to query the the system, query the data. 670 00:36:49,640 --> 00:36:52,680 Yeah, yeah. And and that's, that's, that is 671 00:36:52,680 --> 00:36:53,880 something that we can do as well. 672 00:36:53,880 --> 00:36:57,960 Like you can, you can pop into our account events interface and 673 00:36:57,960 --> 00:37:01,600 sort of ask like, hey, who are the, you know, tell me about the 674 00:37:01,640 --> 00:37:03,840 tell me about the users that have logged in today in terms of 675 00:37:03,840 --> 00:37:06,560 like where did they come from and, and what were they doing? 676 00:37:06,560 --> 00:37:08,160 And is there anything anomalous, right? 677 00:37:08,160 --> 00:37:11,200 Like you can do things like that, but it's, you know, it's 678 00:37:11,200 --> 00:37:14,160 obviously very secured, it's obviously very controlled to 679 00:37:14,160 --> 00:37:18,400 that specific customer data within that customer instance. 680 00:37:18,400 --> 00:37:21,600 And we don't sort of allow, you know, you to sort of go like 681 00:37:21,640 --> 00:37:24,440 allow our system to sort of go out and query the Internet or, 682 00:37:24,440 --> 00:37:26,200 or sort of other external systems. 683 00:37:28,160 --> 00:37:30,080 And the other thing that's interesting too, is that, you 684 00:37:30,080 --> 00:37:34,400 know, we, we have role based access control in the product 685 00:37:35,400 --> 00:37:40,040 and we treat the LLM like any other subset of functionality 686 00:37:40,040 --> 00:37:43,160 within, within the product or any other sort of area that lets 687 00:37:43,160 --> 00:37:45,400 you get access to sensitive information. 688 00:37:45,400 --> 00:37:48,880 So if you don't want certain employees or groups to have 689 00:37:48,880 --> 00:37:52,160 access to the LLM, then you don't give them that right in 690 00:37:52,160 --> 00:37:54,240 the in the admin console and they can't, right. 691 00:37:54,240 --> 00:37:58,640 So if you want to control who sort of has access to those, 692 00:37:58,640 --> 00:38:01,720 that's that's a that's a great way to do it is with RMAC. 693 00:38:02,560 --> 00:38:05,320 That was the question I was going to get into is like how 694 00:38:05,320 --> 00:38:12,400 you control access to either the LLM or what the LLM has access 695 00:38:12,400 --> 00:38:14,880 to. I think that that second 696 00:38:14,960 --> 00:38:17,000 scenario is a lot more difficult, right? 697 00:38:17,160 --> 00:38:20,040 It it is yeah. And so if you think about these 698 00:38:20,040 --> 00:38:23,200 sort of rag based AI systems, they're kind of they're kind of 699 00:38:23,200 --> 00:38:26,160 chains, right? So you, you might, you might 700 00:38:26,160 --> 00:38:29,040 have multiple LLM calls, you might have multiple API calls, 701 00:38:29,040 --> 00:38:31,600 right? You might have a, a, a, a first 702 00:38:31,600 --> 00:38:34,360 date LLM call that's sort of like, Hey, what is this customer 703 00:38:34,360 --> 00:38:37,200 asking for, right? Then you might make a call out 704 00:38:37,200 --> 00:38:40,040 to an API, probably one of my, it's probably going to be one of 705 00:38:40,040 --> 00:38:44,320 our internal APIs to pull data. You might get some context back. 706 00:38:44,480 --> 00:38:48,080 Then you might make another LLM call with that context to get 707 00:38:48,080 --> 00:38:50,080 the final answer, right. So they're, they're implemented 708 00:38:50,080 --> 00:38:53,040 as chains, but that when you implement them as chains like 709 00:38:53,040 --> 00:38:55,280 that, you also have an opportunity to it, you know, 710 00:38:55,280 --> 00:38:58,840 sort of introduce security restrictions at different steps, 711 00:38:58,840 --> 00:39:00,880 right. So it's like, hey, yes, I am not 712 00:39:00,880 --> 00:39:05,400 going to allow any API calls that do a write or, you know, 713 00:39:06,800 --> 00:39:09,920 only can access this subset of APIs within our product. 714 00:39:10,440 --> 00:39:12,880 That that that's kind of the, that's kind of the idea, yeah. 715 00:39:13,280 --> 00:39:18,120 So Steven, where do you see, where do you see this being five 716 00:39:18,120 --> 00:39:21,600 years down the road? Like you said, two years since 717 00:39:21,880 --> 00:39:27,120 Chap GPT kind of hit the mainstream and look where it is 718 00:39:27,120 --> 00:39:29,520 already. I mean, five years might be, I 719 00:39:29,520 --> 00:39:32,240 might as well say, you know, in an eternity. 720 00:39:32,240 --> 00:39:35,560 But where do you see this five years down the road impacting 721 00:39:35,560 --> 00:39:38,360 the identity space? Well, I mean, if I could, if I 722 00:39:38,360 --> 00:39:42,760 could send the, the the shrug emoji to your, to your, your 723 00:39:42,760 --> 00:39:45,680 users, I would. It's it's really hard to say. 724 00:39:45,680 --> 00:39:47,720 I mean, I think I thought you positioned it really well 725 00:39:47,720 --> 00:39:50,560 earlier when you asked like was it was it sort of, you know, was 726 00:39:50,560 --> 00:39:53,640 it sort of a big change on way, the way you think about things 727 00:39:53,640 --> 00:39:56,840 when, when Chad JP take him out because it it, it, it really 728 00:39:56,840 --> 00:39:58,880 was. It's really hard to say because 729 00:39:58,880 --> 00:40:02,360 we're, we're constantly being surprised on what these LMS can 730 00:40:02,360 --> 00:40:04,440 do. I mean, generating like lifelike 731 00:40:04,440 --> 00:40:08,800 video and all kinds of, you know, crazy stuff like this and 732 00:40:08,800 --> 00:40:11,160 that, you know, the AI firms sort of keep bringing out their 733 00:40:11,160 --> 00:40:13,960 latest and, and greatest models, right? 734 00:40:13,960 --> 00:40:16,400 And they just keep out doing themselves. 735 00:40:16,640 --> 00:40:17,920 But I'll tell you one thing that. 736 00:40:18,880 --> 00:40:20,880 That really piqued my interest recently. 737 00:40:21,960 --> 00:40:24,840 It's that the, the large tech firms like, you know, like 738 00:40:24,840 --> 00:40:28,680 Google, Meta, they're, they're, they're seeking out other 739 00:40:28,680 --> 00:40:33,040 sources of, of energy beyond the, the energy grid, right? 740 00:40:33,040 --> 00:40:35,720 So they're, they're talking about like obtaining access to 741 00:40:35,720 --> 00:40:38,920 nuclear energy or building their own nuclear, nuclear energy 742 00:40:38,920 --> 00:40:40,720 infrastructure. That's just insane to me. 743 00:40:40,720 --> 00:40:43,440 But it, it tells me two things, right? 744 00:40:43,480 --> 00:40:48,600 It tells me one, we're taxing the energy grid right now, it's 745 00:40:48,600 --> 00:40:50,960 not going to keep up with the demand of this of this 746 00:40:50,960 --> 00:40:55,520 technology, right? And it's also, it's also a hint 747 00:40:55,520 --> 00:40:59,160 that the those firms are sort of acknowledging that, right? 748 00:41:00,760 --> 00:41:02,760 It means that they're acknowledging that, you know, 749 00:41:02,760 --> 00:41:06,280 any efficiencies that they design into these systems or the 750 00:41:06,280 --> 00:41:09,880 distribution of the compute of these systems, you know, pushing 751 00:41:09,880 --> 00:41:12,800 it down to devices or, or things like that is not going to, it's 752 00:41:12,800 --> 00:41:15,560 not going to keep pace with the demand, right? 753 00:41:15,680 --> 00:41:21,880 So that points, that points to, for me to the technology 754 00:41:21,880 --> 00:41:27,480 reaching a plateau shortly. Because once you run into issues 755 00:41:27,480 --> 00:41:31,600 with resource scarcity, you know, it quickly becomes 756 00:41:31,600 --> 00:41:34,800 geopolitical, right? And you, you know, we, we all 757 00:41:34,800 --> 00:41:37,800 know, we all know humanity has done a great job at sort of 758 00:41:37,800 --> 00:41:41,320 sanely sharing resources and it's in its history, right? 759 00:41:41,920 --> 00:41:46,680 So but yeah, I mean, I kind of think we're only scratching the 760 00:41:46,680 --> 00:41:49,400 surface of this tech right now. I mean, like I say, ChatGPT 761 00:41:49,400 --> 00:41:52,000 came, like you've mentioned, came only out and, you know, two 762 00:41:52,000 --> 00:41:55,000 years ago. Seems like seems like eons. 763 00:41:55,000 --> 00:41:59,280 I mean, that's nothing. That's, you know, I I would 764 00:41:59,280 --> 00:42:03,800 equate unleashing eon the world as, you know, equivalent to the 765 00:42:03,800 --> 00:42:07,240 the advent of the consumer Internet, the invention of the 766 00:42:07,240 --> 00:42:10,480 microchip, you know, the, the industrial revolution. 767 00:42:10,480 --> 00:42:13,120 I mean, I think it has that level of impact on the world. 768 00:42:13,120 --> 00:42:17,240 It's like I I can't imagine where it's going to be in five, 769 00:42:17,240 --> 00:42:20,680 1050 years from now. You know, I mean, it's, it's 770 00:42:20,680 --> 00:42:23,400 insane to think. About mind blowing. 771 00:42:23,400 --> 00:42:28,960 I mean, do you think that it's going to evolve so quickly that 772 00:42:29,280 --> 00:42:35,440 whatever governance is needed to, you know, keep it from 773 00:42:35,440 --> 00:42:41,600 becoming a dangerous weapon, that it's going to outpace that 774 00:42:41,600 --> 00:42:45,760 ability to govern it or that we will be able to govern it? 775 00:42:47,000 --> 00:42:48,720 And it's kind of a crazy question, but. 776 00:42:48,720 --> 00:42:53,760 It's, it's a tough question. I, I, I think it will probably 777 00:42:53,760 --> 00:42:57,960 outpace our ability to, to, to, to govern it, especially with 778 00:42:57,960 --> 00:42:59,880 the fact that, you know, you have sort of open source 779 00:42:59,880 --> 00:43:05,440 capabilities around this. I think, you know, governments 780 00:43:05,440 --> 00:43:08,040 haven't really do done a good job at sort of keeping up with, 781 00:43:08,240 --> 00:43:10,920 with technology governments in the governance in the past. 782 00:43:10,960 --> 00:43:13,680 So there probably will be a crackdown at some point. 783 00:43:13,680 --> 00:43:16,520 But, you know, it just all speaks to more of like keeping, 784 00:43:16,520 --> 00:43:19,680 keeping the keeping the inventors of this type of 785 00:43:19,680 --> 00:43:22,200 technology involved in the, in the governance of this 786 00:43:22,200 --> 00:43:24,440 technology. But it's, it's, it's, it's going 787 00:43:24,440 --> 00:43:26,760 to be a very tough problem. I think it's really going to, I 788 00:43:26,760 --> 00:43:28,920 think we're, it's going to outpace our ability to sort of 789 00:43:28,920 --> 00:43:31,800 keep a, keep a lid on it. Yeah, to get this through as any 790 00:43:31,800 --> 00:43:33,560 guide, it's difficult to keep up with tech. 791 00:43:33,560 --> 00:43:35,960 It just moves too quickly. Yep, Yep, yeah. 792 00:43:35,960 --> 00:43:38,440 And I mean, and our elected officials aren't really aren't 793 00:43:38,440 --> 00:43:42,720 really sort of equipped to, to understand this technology at 794 00:43:42,720 --> 00:43:45,920 any level and, and sort of, you know, the the applications of it 795 00:43:45,920 --> 00:43:47,640 and the impacts of it. They'll get there, they'll get 796 00:43:47,640 --> 00:43:49,800 there. But it's, it's moving so fast 797 00:43:49,920 --> 00:43:52,000 right now, you know? Going back to the power 798 00:43:52,000 --> 00:43:54,720 requirements, sounds like we need a a Tony Stark type person 799 00:43:54,720 --> 00:43:58,160 to come up with an arc reactor that we can stick into a data 800 00:43:58,160 --> 00:44:01,000 center somewhere and have it just power that thing and just 801 00:44:01,000 --> 00:44:03,040 be good. Yeah, I mean, I just, I to me, 802 00:44:03,040 --> 00:44:05,960 it was just insane to see the the tech firms sort of going 803 00:44:05,960 --> 00:44:09,960 after nuclear power to to keep their data centers running. 804 00:44:09,960 --> 00:44:11,640 You know what I mean? It's just crazy. 805 00:44:12,760 --> 00:44:15,800 Microsoft's and, you know, Microsoft AI and you know, 806 00:44:15,800 --> 00:44:19,320 Microsoft nuclear like that is a that is a weird thought to say, 807 00:44:19,480 --> 00:44:22,880 yeah, all right, why don't we go ahead and wrap it up there? 808 00:44:22,880 --> 00:44:25,360 This has been, I mean, we could talk AI all day for sure. 809 00:44:26,240 --> 00:44:29,600 I'm looking at this Trivacity dot AI website and you, you, you 810 00:44:29,600 --> 00:44:31,680 almost stole all the kind of the Thunder of the website because 811 00:44:31,680 --> 00:44:33,960 on the screen I'm looking at right now, you know, it's, it 812 00:44:33,960 --> 00:44:38,440 has like this little picture of access control, you know, 813 00:44:38,560 --> 00:44:40,480 settings and you got the AI assistance. 814 00:44:40,480 --> 00:44:42,960 I, you know, basically asking, hey, are these settings secure? 815 00:44:42,960 --> 00:44:45,800 Which I think is a really cool way to approach securing the 816 00:44:45,800 --> 00:44:47,280 product, right? Especially you mentioned, you 817 00:44:47,280 --> 00:44:48,840 know, being able to train it on your documentation. 818 00:44:48,840 --> 00:44:51,400 It's like, OK, here's exactly what I'm looking for. 819 00:44:51,920 --> 00:44:55,800 And I love that idea of leveraging those AI assistants 820 00:44:55,800 --> 00:44:58,240 to help with that because I don't want to pick up the phone 821 00:44:58,240 --> 00:45:00,680 and talk to somebody. I want to do it myself, right? 822 00:45:00,680 --> 00:45:03,280 I want to be able to look it up and say, OK, click this, do this 823 00:45:03,280 --> 00:45:04,040 here. OK, good. 824 00:45:04,040 --> 00:45:06,400 Now I kind of get it. So it's a really cool example of 825 00:45:06,400 --> 00:45:07,640 how how you guys are approaching it. 826 00:45:07,840 --> 00:45:11,360 You've got it. So let's talk a little bit about 827 00:45:11,640 --> 00:45:17,800 astrophotography because I, I did my first astrophotography 828 00:45:17,800 --> 00:45:19,440 just a few weeks ago. Oh, yeah. 829 00:45:19,760 --> 00:45:25,440 So I'm in the Asheville, NC area and we had Northern Lights 830 00:45:25,440 --> 00:45:27,720 basically come down as far down as we were. 831 00:45:27,720 --> 00:45:29,600 And that's the first time I've ever experienced anything like 832 00:45:29,600 --> 00:45:32,320 that. And it was almost invisible to, 833 00:45:32,680 --> 00:45:34,920 you know, the naked eye. But when I pulled out my phone 834 00:45:34,920 --> 00:45:37,400 and that's all I had to kind of take pictures. 835 00:45:37,920 --> 00:45:41,080 All of a sudden I've got this picture and it's like Reds and 836 00:45:41,080 --> 00:45:44,600 purples are in the sky and it's, you know, 2 in the morning or 837 00:45:44,600 --> 00:45:45,960 whatever it is. I'm like, wait a second. 838 00:45:46,160 --> 00:45:49,360 I like, that's not what I saw, but that's what the camera 839 00:45:49,360 --> 00:45:51,280 picked up. And I'm just fascinated by this 840 00:45:51,280 --> 00:45:52,440 idea. So I started looking at it like, 841 00:45:52,440 --> 00:45:55,120 well, can I download an app or something to kind of, you know, 842 00:45:55,200 --> 00:45:58,360 capture this better than, than just the, you know, the iPhone? 843 00:45:58,840 --> 00:46:01,080 And so I'm curious because I think you're into the 844 00:46:01,080 --> 00:46:03,440 astrophotography stuff too. I I want to get into this a 845 00:46:03,440 --> 00:46:05,320 little bit more with you. Yeah, yeah. 846 00:46:06,280 --> 00:46:08,080 So, you know, it's, it's actually something that I 847 00:46:08,080 --> 00:46:10,800 recently got into as well, like as in this year. 848 00:46:11,840 --> 00:46:14,200 And it's, you know, it's something I've I've I've wanted 849 00:46:14,200 --> 00:46:16,920 to, I've had a I've had a telescope for a while, like a 850 00:46:16,920 --> 00:46:20,040 nice one, but I've never gotten into sort of the like post 851 00:46:20,040 --> 00:46:24,520 processing part of it, you know. And so I think one of the things 852 00:46:24,520 --> 00:46:29,680 that was sort of holding me up on it was, you know, some of 853 00:46:29,680 --> 00:46:32,640 these rigs that you build for sort of capturing deep space 854 00:46:32,640 --> 00:46:35,720 objects can be like incredibly expensive, like, you know, five, 855 00:46:35,720 --> 00:46:38,560 $10,000 if you really want to build like a really nice rig, 856 00:46:38,920 --> 00:46:41,760 But there's actually this really cool and something you might 857 00:46:41,760 --> 00:46:43,160 want to look into if you have an interest. 858 00:46:43,160 --> 00:46:46,440 There's really there's these really cool set of like smart 859 00:46:46,440 --> 00:46:48,960 telescopes coming out where you can. 860 00:46:49,320 --> 00:46:53,160 They're not very big. They, you can sort of put them 861 00:46:53,160 --> 00:46:57,560 out, you know, in your yard and, and train them on a, on a, on an 862 00:46:57,560 --> 00:46:59,800 object that you want to record. And then they have an app where 863 00:46:59,800 --> 00:47:02,480 you sort of control it and, and capture images. 864 00:47:02,480 --> 00:47:05,400 And they actually can produce some pretty amazing images. 865 00:47:06,040 --> 00:47:08,840 So they'll take exposures, like they'll take like 10 second 866 00:47:08,840 --> 00:47:11,600 exposures and you know, you leave them out there and maybe 867 00:47:11,600 --> 00:47:16,240 over the night you take 510,000, you know, 10 second exposures. 868 00:47:16,520 --> 00:47:19,000 And then they have stacking tools, right? 869 00:47:19,000 --> 00:47:22,800 So you take these, all of these images, you stack them together. 870 00:47:23,200 --> 00:47:25,960 You know, it does some crazy calculations on the pixels to 871 00:47:25,960 --> 00:47:30,080 decide what is noise and what is, what is image, you know, 872 00:47:30,080 --> 00:47:34,320 what is, what is signal And, and you get this beautiful image of 873 00:47:34,320 --> 00:47:36,960 like a deep space object, right? Like, so like of a, you know, of 874 00:47:36,960 --> 00:47:41,240 a nebula or you know, of a, of a Galaxy, right. 875 00:47:41,560 --> 00:47:45,280 And I, I just, I was, like I said, I was afraid from the cost 876 00:47:45,280 --> 00:47:48,320 requirements and I was afraid that I wouldn't like the post 877 00:47:48,320 --> 00:47:51,160 processing angle of it. Just, you know, because it's, 878 00:47:51,160 --> 00:47:54,640 it's, it's a whole new sort of technology to learn, but I, I 879 00:47:54,640 --> 00:47:57,320 love it. It's, it's actually another 880 00:47:57,320 --> 00:48:02,280 place that the AI has made an appearance because you can, they 881 00:48:02,280 --> 00:48:06,760 have these sort of AI driven noise algorithms that are that, 882 00:48:06,760 --> 00:48:09,400 that people have developed where they try to figure out like what 883 00:48:09,400 --> 00:48:12,160 is signal and what is noise and sort of remove the noise from an 884 00:48:12,160 --> 00:48:13,960 image. So you can, you know, you can 885 00:48:13,960 --> 00:48:17,720 produce these visually brilliant images from a little telescope 886 00:48:17,720 --> 00:48:20,000 sitting out on your backyard because of, because of the way 887 00:48:20,000 --> 00:48:22,360 you're stacking the images and the way that you're, you're, 888 00:48:22,360 --> 00:48:27,360 you're denoising them, right. So yeah, it's, I've actually, I, 889 00:48:27,520 --> 00:48:29,520 you know, I've spent like it's been a really, we've had a 890 00:48:29,520 --> 00:48:31,880 really run good run of the sort of clear nights here in 891 00:48:31,880 --> 00:48:33,360 Virginia. And I've been out, I've been out 892 00:48:33,360 --> 00:48:36,280 in my backyard like almost every night doing, you know, taking 893 00:48:36,280 --> 00:48:39,320 pictures, so. Well, that's the key, right, is 894 00:48:39,320 --> 00:48:41,360 light pollution kind of really impacts that. 895 00:48:41,360 --> 00:48:44,560 So if you're in a in a rather remote area where there's less 896 00:48:44,560 --> 00:48:47,600 of the light pollution you're, you're more prone to be having 897 00:48:47,760 --> 00:48:50,240 to have better quality for the images, right? 898 00:48:50,280 --> 00:48:51,960 Yeah. And they they call that the, 899 00:48:52,600 --> 00:48:56,760 they call that the Bortle scale. And so it goes from, I think it 900 00:48:56,760 --> 00:48:58,720 goes from like 1:00 to 9:00, I think. 901 00:48:58,720 --> 00:49:03,600 But I'm in a Bortle 5 area. So I'm right in the middle 902 00:49:03,600 --> 00:49:06,320 because I'm right outside of a, of a major airport where I live. 903 00:49:06,320 --> 00:49:09,280 So that you get a lot of light pollution from, but we actually, 904 00:49:09,280 --> 00:49:12,520 we go down to the Outer Banks. I'm North Carolina a lot I'm 905 00:49:12,560 --> 00:49:15,040 sure you're familiar with. And a lot of people don't 906 00:49:15,040 --> 00:49:18,200 realize that is actually one of especially the Ocracoke areas, 907 00:49:18,200 --> 00:49:20,280 one of the darkest areas in the United States. 908 00:49:20,280 --> 00:49:22,880 And it is one of the darkest areas on the, on the East Coast 909 00:49:22,880 --> 00:49:24,240 as well. So you can actually get Bortle 910 00:49:24,240 --> 00:49:27,400 2, which is the second darkest on the Bortle scale. 911 00:49:27,880 --> 00:49:31,400 So I and it having the little smart scope, you can, you can 912 00:49:31,400 --> 00:49:33,560 kind of take it down there. It's not, it doesn't require a 913 00:49:33,560 --> 00:49:35,480 lot of space. So like when we go down there, 914 00:49:35,480 --> 00:49:37,720 I'll, I'll, I'll take it down there and sort of put it out on 915 00:49:37,720 --> 00:49:39,920 the back deck and, and get some really good pictures because 916 00:49:39,920 --> 00:49:42,360 it's so dark down there. So from what I understand of 917 00:49:42,360 --> 00:49:46,880 Ocracoke Island, you take a ferry to get there you have to 918 00:49:46,880 --> 00:49:50,840 have a 4 by 4. You drive on the beach and pick 919 00:49:50,840 --> 00:49:54,200 your spot and you can camp or whatever. 920 00:49:54,200 --> 00:49:59,160 A lot of people go surf fishing. So you go onto this island and 921 00:49:59,160 --> 00:50:01,880 you set up your telescope and your camera. 922 00:50:02,200 --> 00:50:03,560 Yeah, Yeah, you can do that, Yeah. 923 00:50:03,560 --> 00:50:06,040 I mean, Ocracoke, you have to take a ferry to Yeah, but there 924 00:50:06,040 --> 00:50:09,760 actually is a Route 12 goes. It's a, you know, a major road 925 00:50:09,760 --> 00:50:11,480 goes all the way down to the city of Ocracoke. 926 00:50:11,480 --> 00:50:15,680 But yes, if you go out on the beaches of Ocracoke when it's a 927 00:50:15,680 --> 00:50:20,680 clear night, you can see the, the Milky Way without any 928 00:50:20,680 --> 00:50:23,360 imaging equipment at all. And it's it's amazing. 929 00:50:23,680 --> 00:50:25,280 Ocracoke's beautiful. If you haven't been there, I 930 00:50:25,280 --> 00:50:27,600 highly recommend it. So that's. 931 00:50:27,840 --> 00:50:29,560 That's what our ancestors did, right? 932 00:50:29,560 --> 00:50:32,440 They just looked up in the sky. There's no light pollution. 933 00:50:32,600 --> 00:50:36,000 Yeah. Yeah, We did a episode pretty 934 00:50:36,000 --> 00:50:43,320 recently with a guy named Nitin. He's the CTO over at Civil 935 00:50:43,320 --> 00:50:49,640 Security and he rents part. He's in a group of people who 936 00:50:49,640 --> 00:50:55,560 rent AA telescope and Jeff made the the statement, which I 937 00:50:55,560 --> 00:50:57,720 thought was spot on. Like there's supposed to be an 938 00:50:57,720 --> 00:51:01,080 expensive telescope if it takes several people to kind of pull 939 00:51:01,080 --> 00:51:03,000 their money. To bring a trailer and 940 00:51:03,000 --> 00:51:05,440 everything. Oh my gosh, Yeah, yeah, yeah. 941 00:51:05,440 --> 00:51:09,920 Some of some of the, the, the Dobsonian telescopes are really 942 00:51:09,920 --> 00:51:10,840 big. Yeah. 943 00:51:10,840 --> 00:51:13,440 So they're, and they're actually less, less expensive. 944 00:51:13,760 --> 00:51:16,560 I don't know why that is, but but yeah, they can be pretty 945 00:51:16,560 --> 00:51:18,520 big. But yeah, one of the reasons I 946 00:51:18,520 --> 00:51:21,600 got into the sort of smart telescope area was because it's 947 00:51:21,600 --> 00:51:23,000 they're portable. You know what I mean? 948 00:51:23,000 --> 00:51:26,000 You can, it can, it can pack down into a little box and I can 949 00:51:26,000 --> 00:51:28,640 just throw it in the back of my truck on my way down, you know, 950 00:51:28,640 --> 00:51:30,760 to the Outer Banks. So I had. 951 00:51:30,760 --> 00:51:33,040 To look up the Bortle scale because I hadn't heard of that 952 00:51:33,040 --> 00:51:35,080 before, because I know I live in an area that's got less Royce 953 00:51:35,080 --> 00:51:37,480 provision and it looks like the Asheville area is around a three 954 00:51:37,480 --> 00:51:38,560 or a four. Wow. 955 00:51:39,000 --> 00:51:39,880 Yeah, pretty. Good. 956 00:51:39,960 --> 00:51:42,600 And I've got a mountain range between where my house is and 957 00:51:42,600 --> 00:51:45,160 the rest of Asheville, so it's even darker, so I feel like I'm 958 00:51:45,160 --> 00:51:47,160 in a good spot. So if I, yeah, if I drive a 959 00:51:47,160 --> 00:51:50,400 little bit West of where I am, I can get to the Shenandoah 960 00:51:50,400 --> 00:51:53,680 Mountains and I can get some really good, really good dark, 961 00:51:53,960 --> 00:51:57,120 dark spots out there. But, but again, like even in a, 962 00:51:57,120 --> 00:51:59,320 even in a highlight pollution area, you can pull a lot of the 963 00:51:59,320 --> 00:52:01,960 light pollution out of the images in the post processing 964 00:52:01,960 --> 00:52:04,760 part of of this. So there's people that take 965 00:52:04,760 --> 00:52:08,160 great pictures like out of Tokyo where it's portable 9, you know, 966 00:52:08,160 --> 00:52:12,360 So it's, you know, some of the denoising algorithms are, are 967 00:52:12,360 --> 00:52:14,920 pretty fascinating. You stack enough images you can 968 00:52:14,920 --> 00:52:17,400 you can figure out what's what's noise, what's light pollution 969 00:52:17,400 --> 00:52:18,840 and what's what's image you know. 970 00:52:19,840 --> 00:52:22,320 Of course you get those cool like motion ones across the sky 971 00:52:22,320 --> 00:52:23,760 right where it's. Like yeah, with the starting. 972 00:52:23,760 --> 00:52:25,440 Image and here's the only thing that's moved, right? 973 00:52:25,440 --> 00:52:26,240 Yeah. And stuff like that. 974 00:52:26,240 --> 00:52:28,480 Yeah. So if someone wanted to get into 975 00:52:28,480 --> 00:52:30,560 this, what would be the beginner? 976 00:52:30,920 --> 00:52:35,120 And let's say it's me on the on my back deck trying to have 977 00:52:35,120 --> 00:52:38,800 maybe just a slightly better set up to do some astrophotography. 978 00:52:39,440 --> 00:52:45,000 So, so the, the, there's a company called CWO and they have 979 00:52:45,000 --> 00:52:49,200 a, they have a couple of scopes called the sea stars. 980 00:52:49,200 --> 00:52:52,240 And so there's like AC * S 30 and AC star S50. 981 00:52:52,560 --> 00:52:57,280 And they are the, the, the highest quality, sort of easiest 982 00:52:57,280 --> 00:52:59,360 beginner scopes that, that you can find. 983 00:53:00,040 --> 00:53:03,600 They'll actually livestack with, they'll livestack the, they can 984 00:53:03,600 --> 00:53:06,960 livestack the, the images that you take while you're using the 985 00:53:06,960 --> 00:53:08,640 app. So you don't even really need to 986 00:53:08,640 --> 00:53:11,000 do post processing on your own, unless you want to throw 987 00:53:11,000 --> 00:53:12,800 something into like Photoshop or something. 988 00:53:13,360 --> 00:53:17,000 They'll do the stacking for you. So you pop the scope out in the, 989 00:53:17,000 --> 00:53:19,760 in the, you know, wherever you want to, you want an image, you 990 00:53:19,760 --> 00:53:22,920 tell it, you actually can tell it, like what deep space object 991 00:53:22,920 --> 00:53:26,200 or what the moon or, or what planet you want to look at. 992 00:53:26,200 --> 00:53:28,480 It'll point the scope at that, it'll start taking pictures, 993 00:53:28,480 --> 00:53:30,880 it'll stack it for you and you get a nice little image on your 994 00:53:30,880 --> 00:53:35,440 phone and your, or your tablet. And so I'm a big fan, big fan of 995 00:53:35,440 --> 00:53:37,520 that company. What do you do with the pictures 996 00:53:37,520 --> 00:53:41,560 that you take? I post them on Instagram and and 997 00:53:41,720 --> 00:53:45,000 and Facebook and, you know, share them with share them with 998 00:53:45,000 --> 00:53:46,040 my buddies. Wallpapers. 999 00:53:46,040 --> 00:53:47,400 Print them out. Put them on behind you. 1000 00:53:48,080 --> 00:53:50,520 Yeah, well, maybe I haven't done that yet, but that that would be 1001 00:53:50,520 --> 00:53:53,560 a good idea. You know, my kids also enjoy 1002 00:53:53,760 --> 00:53:57,280 seeing what I'm doing and taking pictures off, so I share it with 1003 00:53:57,280 --> 00:54:00,480 them too. Well, maybe Sea Star wants to 1004 00:54:00,480 --> 00:54:01,600 sponsor an episode in the future. 1005 00:54:02,080 --> 00:54:03,440 There you go. Yeah, stuff like that. 1006 00:54:03,520 --> 00:54:06,080 They'll appreciate the mention. It's a tiny little. 1007 00:54:06,080 --> 00:54:08,080 I'm looking at the one right now and it's definitely a tiny 1008 00:54:08,080 --> 00:54:09,440 little thing. It's much smaller than I 1009 00:54:09,440 --> 00:54:10,720 anticipated it being. Yeah. 1010 00:54:10,760 --> 00:54:14,120 Well, that's super cool. Which I think is probably a good 1011 00:54:14,120 --> 00:54:17,440 spot to leave it with a product mentioned for another company. 1012 00:54:17,800 --> 00:54:22,760 So let's get back to stravacity. Stravacity dot AI Steven, you, 1013 00:54:22,800 --> 00:54:24,560 you've been great again. You know, I hope we can talk 1014 00:54:24,560 --> 00:54:28,160 again in the near future. I really like what you guys have 1015 00:54:28,160 --> 00:54:30,960 done with the the AI assist. And so I'm definitely encourage 1016 00:54:30,960 --> 00:54:33,960 people to go check it out. And I've been a big fan of what 1017 00:54:33,960 --> 00:54:35,560 you guys been working on for the last, you know, couple years 1018 00:54:35,560 --> 00:54:36,760 now. So it's been exciting to see 1019 00:54:36,760 --> 00:54:40,120 that evolution, you know, happen over and I won't say it real 1020 00:54:40,120 --> 00:54:42,960 time because, you know, a couple years is is a blink in the eye 1021 00:54:42,960 --> 00:54:45,280 of a tech spot, but it's been really impressive. 1022 00:54:46,040 --> 00:54:48,680 Thanks, I and I and I always appreciate talking to you guys. 1023 00:54:48,680 --> 00:54:51,600 I hear, I hear that you guys are like celebrities at conferences 1024 00:54:51,600 --> 00:54:54,600 now and people walk up to you and recognize you and stuff. 1025 00:54:54,600 --> 00:54:57,720 So I'm, I'm, I'm honored that you, you would have me on your 1026 00:54:57,760 --> 00:55:00,320 on your show again. So I'm looking forward to the 1027 00:55:00,320 --> 00:55:02,280 next time. Well, we appreciate that. 1028 00:55:02,480 --> 00:55:05,040 Yeah, for sure. And definitely appreciate all 1029 00:55:05,040 --> 00:55:08,760 the wishers to come up and and I try not to be awkward when I can 1030 00:55:08,760 --> 00:55:11,640 do it. All right, let's go and leave it 1031 00:55:11,640 --> 00:55:14,920 there for this week. You can find Strabasi on the web 1032 00:55:14,920 --> 00:55:17,000 again, Strabasi dot AI. We'll have links in our show 1033 00:55:17,000 --> 00:55:18,800 notes and people can check that out. 1034 00:55:18,800 --> 00:55:21,600 And then Steven, I'll put a link to your LinkedIn profile as well 1035 00:55:22,080 --> 00:55:24,440 in case people have questions or maybe just want to talk about 1036 00:55:24,440 --> 00:55:26,400 astrophotography. I'm sure, I'm sure you would 1037 00:55:26,400 --> 00:55:27,600 appreciate either of those things. 1038 00:55:28,960 --> 00:55:34,480 And as far as we go, we're on the web, idacpodcast.com and we 1039 00:55:34,480 --> 00:55:37,080 got the YouTube channel going now, idacpodcast.tv. 1040 00:55:37,080 --> 00:55:38,400 So feel free to like and subscribe. 1041 00:55:38,400 --> 00:55:40,080 Jim's shaking his head. Like make sure I get that plug 1042 00:55:40,080 --> 00:55:42,280 out. But yeah, thanks for everyone 1043 00:55:42,280 --> 00:55:45,800 for listening and are watching and we'll talk with you all in 1044 00:55:45,800 --> 00:55:49,640 the next one. You've been listening to 1045 00:55:49,640 --> 00:55:53,560 Identity at the Center. We hope you've enjoyed the show. 1046 00:55:53,760 --> 00:55:57,880 Make sure to like, rate and review, and we'll be back soon. 1047 00:55:58,120 --> 00:56:00,400 But in the meantime, hit the website at 1048 00:56:00,400 --> 00:56:06,760 identity@thecenter.com. See you next time on Identity at 1049 00:56:06,760 --> 00:56:07,680 the Center.