Latest / NC Tweener Talks / Igor Jablokov: From a cave to IBM to Amazon/Alexa, then $160m+ at Pryon

Transcript

Igor Jablokov: From a cave to IBM to Amazon/Alexa, then $160m+ at Pryon · NC Tweener Talks

• 00:00:03 - Igor Jablokov I always tell people, you know, founders don't need a compass. You know, it's just, we just go out there and we just survive on, on Spidey sense and intuition and stuff like that as well. That's where we're good at. It's like it doesn't bother me to have two engines out, you know, with bird strike a la Captain Sully, it's like, I'll find you a place to land. Welcome to Triangle Tweener Talk, a weekly podcast by Builders for Builders where we explore the startup journey from the idea to the exit and all the lessons in between. With an exclusive focus on founders from the Triangle region of North Carolina. Tweener Talks is produced by Earfluence. Now here is your host, serial founder and General partner of the Triangle Tweener Fund, Scott Wingo. 00:00:52 - Scot Wingo Welcome to this episode of Triangle Tweener Talks featuring Igor Yablokov. This episode is brought to you by our sponsors. Extensus HR a professional employer organization known as a P.E. extensus HR empowers tech founders and growing businesses to scale smarter. They take HR administration off your plate, managing payroll recruiting, employee benefits and retirement plans, compliance risk and more so you can focus on growing your business and innovating. 00:01:19 - Scot Wingo For over 25 years, they've provided the tools needed to stay competitive in today's market. Eisner Amper One of the world's largest business consulting firms, they have a dedicated technology practice offering outsourcing, accounting, tax and advisory services. Their experienced professionals serve more than 2,000 technology companies from early stage startups to public enterprises. Bank of America whose transformative technology group helps game changing tech businesses and founders realize their boldest ambitions across a wide range of technology sectors. Aligned Technology Group Unlock the power of the cloud with Aligned Technology Group's free Catalyst program. As an AWS Advanced Tier Consulting partner, they specialize in cost optimization, security assessments and tailored cloud strategies for startups and SMBs. 00:02:02 - Scot Wingo Robinson Bradshaw a full service business law firm with a passion for supporting the Triangles of startup ecosystem. Whitley Recruiting Partners they specialize in recruiting top tech talent for growth stage startups in the Triangle. They target industry specific entrepreneurial employers employees who drive immediate results to a fast, accurate and scalable recruiting process. Smashing boxes, a Durham based design centric digital transformation. And last but not least, this podcast would not be possible if it weren't for our awesome partners at Earfluence. 00:02:32 - Scot Wingo If you're interested in doing anything around podcasting, give them a shout@earfluence.com let's talk about Ibor Yablikov I known Igor for a while, and when we get in this interview, one of the fun things I find doing this is I learned stuff about people I didn't know. This was a. This one starts with a pretty big surprise. This story starts in Greece, which, which I had kind of guessed, but in Greece we are going to zoom in and there's a tiny Greek island. 00:03:00 - Scot Wingo On that island is a little cave, and that cave overlooks a lagoon. And in that cave is a young lad named Igor who would go on to be one of the top AI thinkers not only in the Triangle, but in the world. His latest company, Prion, is set to be the Triangle's next unicorn and has already raised over $160 million. In this story, we follow Igor's journey from that little cave in Greece to Philadelphia, then to getting an engineering degree and an mba. 00:03:29 - Scot Wingo After that, he goes to IBM where he is serendipitously put on this big project called AI. He quickly leads teams and does some really innovative stuff. Then in 2006, he and his brother decide to start their own company called Yap. And that is the first talk to tech solution. I'm not going to spoil what happens with Yap, but it would be, it would not be crazy to say you have probably used this technology hundreds, if not thousands of times. 00:04:02 - Scot Wingo Ten years ago I had a coffee with Igor and he was, this was after, before he had started Prion. And he basically said, I was like, what do you, you know, what's interesting to you? He said, oh, the AI future is going to be amazing. They're gonna, you know, we're gonna have these little, you know, AI minds in, in the, in the cloud, and they're going to be able to talk to databases and, and learn more about your business than you could ever possibly know than any analyst or team of anal. 00:04:29 - Scot Wingo And then we're going to be able to put them in these little agents and they're going to run around and do things on our behalf. It's going to be like nothing we can ever imagine. Here we are, 10, 11 years later, and this is coming to fruition. So on this pod, we're going to dig into that story of the little fella in Greece and how he gets from the Cave to basically being a top mind in AI. 00:04:53 - Scot Wingo We're going to go over what happened with Yap. That's a great story. And then how he came up with Prion and get an update for what's going on there. And then at the end, maybe most importantly, we're going to pick Igor's brain on what he thinks is going to happen in the next 10 years. So he's obviously got a good, good track record of predicting things. So it's going to be interesting to see if this comes true or not. I hope you enjoy this conversation with Igor as much as I did. 00:05:17 - Igor Jablokov I was actually born in Greece to two artist parents, and we lived in a cave and a Santorini kind of a thing. Yeah, a little island. No running water, no tv, no radio, no electricity. And it was rather idyllic. I mean, we were open right to a lagoon. And while some of you had cats and dogs as pets growing up, you know, I just remember these dolphins coming into the lagoon and there were turtles and seahorses and things of that sort. It was a pretty good place to start. And I would have said that, you know, one of the reasons why, you know, maybe I was interested in natural language and things of that sort is I once spotted a hurt dolphin that was struck by a propeller blade, you know, similar to the lot of manatees, you know, down in southern Florida. And that's when the first idea of why can't I talk to you? Popped in my head, you know, then a few years. 00:06:17 - Scot Wingo Do you think dolphins are using language, like a relatively sophisticated language? 00:06:21 - Igor Jablokov Oh, yeah, yeah, for sure. 00:06:22 - Scot Wingo And then do you think LLMs will help us? AI will help us decode that one day? 00:06:27 - Igor Jablokov Yeah, one day. And I think it's in the same way that people used to come to new nation states and consider the individuals that they found there, savages. They didn't understand their culture, their religion, their technologies and things of that sort. But then once we had the rise of machine translation, things of that sort, it allowed us to better relay to one another. I think on the day that we can talk to animals and plants, you know, we'll. We'll see that the humans that existed before that time and after that time are just missing a rich experience, you know, on this planet. So I'm looking forward to that. Yeah. 00:07:07 - Scot Wingo I'm going to hazard a guess you've read the Hitchhiker's Guide to the Galaxy, which has. The dolphins are much smarter than us and is one of its fun. 00:07:15 - Igor Jablokov Although I got to tell you, as you know, think about this, if I spend my life's journey trying to talk to dolphins, I already know what they're going to tell me. 10, 20, 30, 40, 50 years from now that I'm going to be in front of them and saying, hey, I spent my entire life trying to talk to you. And things of that sort. And they're just going to talk back, you should have been swimming. So I already know that they're going to be snarky, that I should have not been so pasty, stuck inside of buildings and stuff like that, huddled over a computer and things of that sort. Funny enough, my mother then had this inclination and thought that, hey, I want you to be part of the computer age. And so at the age of six, she moved us to the States. We moved to Philadelphia to live with my grandparents who were there at the time. And, and that's how I got stateside. 00:08:10 - Scot Wingo Okay, so you're six years old and you moved from a cave in Greece to Philadelphia. That's a pretty. And I guess that's a big change for a six year old. 00:08:18 - Igor Jablokov Yeah. I got to tell you, when I landed in JFK and drove through New York City, I was looking up at all of these buildings and things of that sort. And it felt like that first scene in Blade Runner, you know, where they're flying around all the buildings and it was all that fog and stuff like that. I felt like I landed on a, on a different planet. Yeah, I mean, it was just completely, completely wild to me in terms of just this stark contrast from going some, you know, starting somewhere so rustic, of course, and just nature to this technological marvel. 00:08:53 - Scot Wingo Yeah. Okay, so you're in Philly and then, you know, your mom wants you to be part of the computer age. I'm imagining a computer here pretty soon. 00:09:01 - Igor Jablokov Yeah, yeah. She puts a tandy color computer in front of me and I start just start tapping away at it. It had only 4k of memory and I remember it had a tape drive. It kept eating my basic programs. I could never get it to like save C, save this and C load that. You know, it's sort of like that kid, you know, fighting the furnace in the home alone. You know, I could never, never get it to work. 00:09:28 - Scot Wingo Yes. So then what? For folks that don't enjoy this, I had a TI99.4A, which also had a tape. And it was a cassette. Yeah, like a lot of people have seen the, the cassettes. So what it would do is it was basically paying a modem. So. So if you ever listen to the cassette, it's like it was as if you're listening to them. So it just like saved the data in an analog and then, then, yeah, and then it was very, very easy for them to eat your stuff. So you do all this work on that little cassette and then like it would, you know, it would Pull out the tape into the machine and you'd pull out the tape and it would just be. Sometimes it'd be crinkled beyond belief. Totally ruined. 00:10:00 - Igor Jablokov Oh yeah, yeah. And there was no Internet, no email, no web browsers. It was all character based and things of that sort. And you're just writing your little programs. You have little cartridges you can plug into it to play some early games and things of that sort. It wasn't until years later that we started getting modems and hitting BBSs and stuff like that. 00:10:21 - Scot Wingo I find one of the things that taught me though is to have a pretty good memory of stuff because if it's going to get destroyed, I need to keep the. Yeah, I need to keep it in my head. So when you're pretty young and you're starting to do these memory exercises, it can benefit you later in life. 00:10:33 - Igor Jablokov Yeah, well, I mean, and it becomes. We don't realize. It's sort of like in Raleigh, we lost power here in the last few days. Right. Because of the ice storms and things of that sort. And it's just that quiet that ends up happening when everybody loses power. And you don't realize how much of that din is happening behind the scenes as well. I think that's the one way that you can describe childhood in the 80s, relevant to childhood now is we had that calm and that quiet which allows your mind to wander in some ways. And you got to read books and comic books and go to actual movie theaters and things of that sort. Similar to anybody that's going to school. You always had a conflict with different individuals and things like that. But things moved at a slower pace where you can think through that and, and figure out how to debate people and then come back and, and you know, create, you know, communities of interest and things of that sort. It's, it's a lot, it's a double edged sword now. You have a lot more access to things, but at the same time you don't have that ability to think through these different activities. 00:11:44 - Scot Wingo Yeah, it makes it very hard to have a long attention span. 00:11:46 - Igor Jablokov Yeah, yeah, yeah. 00:11:47 - Scot Wingo I fight that every day. So you're six in Philly. How did you, you know, walk us through how you got to college? 00:11:55 - Igor Jablokov Yeah, I mean they, they put me in front of that computer because I didn't speak English. Right. So I spoke Greek. So it took me a while, you know, to learn English. And then my mother ended up getting remarried so we moved up to, to Montreal. And I remember the first day that we saw just the town buried in, in Snow. And my brother and I were really, really excited, exclaiming to our stepfather about the snow day that we were about to have. And he's like, what's a snow day? And we said, oh, we're going to stay home, right? We get to play in the snow. He's like, no, otherwise you'd never go to school. And we're like, you got to be kidding me. And so again, keeping with the Star wars references, they bundles us up like Hoth, the ice planet and they just pushed us out the front and we're going through all these snow drifts, taking public transportation. And the Quebecois, the French Canadians, they had a sense of humor because for first period then you had to disrobe and jump into a freezing cold Olympic sized swimming pool. For phys ed, you know, being first period, it's like, as it is, you're already like, what are, you know, what is all this? And I have to survive inside of a tauntaun. And they're throwing you, you know, as a polar bear into that as well. And so I have to say Montreal is a, you know, one of the coolest cities on, on planet Earth because there's so many cultures that are converging there and it has that edge, it has a little bit of an edge that they don't have in Toronto. Toronto is a little bit more sterile and it's pretty, a pretty cool place to, you know, to grow up. And then I returned, returned back to the States to finish up high school and that's when I ended up in undergrad. I started as a nuclear engineering candidate at Penn State originally, but then that was coming easy to me. And then it wasn't hard enough. It wasn't, it wasn't hard enough. And then also I knew that there were no jobs, you know, for, for folks with that background. And then I switched over to computer engineering because I knew computers were going to be in the DNA of everything and I can work in any industry I wanted. 00:14:11 - Scot Wingo Why were you, why did you start a nuclear, Were you interested in like, was there something that kind of intrigued you about it? Like submarines or. 00:14:18 - Igor Jablokov Yeah, the submarine stuff was interesting. The, the using the small nuclear engines for spacecraft was interesting as well. But you know, I was, you know, also a nerd that, you know, in the, in the cafeteria in high school, I would be designing fusion reactors, you know, sketching them out. They looked like donuts. And then also, you know, reading the ubiquitous atom. So there was something really interesting about all, all manners of like green energy and like high powered, you Know, generating a lot of electricity from, from, you know, fundamental metals and stuff like that was super cool. 00:14:55 - Scot Wingo Do you ever get your hands on some. I know a fair number of nerdy high schoolers that had a quest to get some uranium. Did you ever get your hands on any uranium? 00:15:02 - Igor Jablokov I have gotten my hands on some things and some. I did work on NASA project at Penn State and they're afraid of us doing that stuff on campus. So we had a building off campus where we would be doing our things. 00:15:18 - Scot Wingo Yeah, interesting. Okay. 00:15:19 - Igor Jablokov And they had an on campus reactor and things like that. I worked also for a Navy lab there where they were doing a lot of the submarine hull designs, torpedo designs and things of that sort, which was super cool. 00:15:32 - Scot Wingo Cool. So you switched from nuclear to computer, where all the hip kids are. And then did you stay at Penn State and finish that or. 00:15:40 - Igor Jablokov Yeah, yeah, finish, finish that. And then. And then during that time I was getting internships at IBM and then eventually joined that company full time because I thought it was super cool. That had a global presence. They had the systems view of technology. Instead of just working on networking, just working on software, just working on chips, just working on servers, they essentially had their mitts and everything. And that was, you know, personally interesting to me. 00:16:08 - Scot Wingo Yeah, kind of full stack, kind of vertical approach versus like up in Schenectady or like in the. 00:16:14 - Igor Jablokov No, actually my. For my first job, that's how I ended up in the Carolinas. 00:16:18 - Scot Wingo Okay. So it's nice. 00:16:19 - Igor Jablokov Yeah. So I moved, moved down to Charlotte because there used to be a relatively large microelectronics plant down near uncc. That's where I started my career as a research engineer at IBM Microelectronics. I remember, I think one of the first weeks that I started, they couldn't find me. You know, I wasn't, I wasn't in my office and things of that sort. And eventually they saw me roaming around the actual manufacturing floor and I'm like, how am I going to design things if I don't know how people put them together? And I want to talk to the actual workers that are putting these things together so that I can get their real life, you know, guidance and things of that sort. 00:17:00 - Scot Wingo This is like a fab, like silicon fab type of situation. Was this like micrometer kind of at that point? Probably. 00:17:08 - Igor Jablokov Yeah, there were. No, but they were building things like video cards for Intergraph. They were actually building some of the motherboards for Radius when Apple for a time was licensing out its os. So they were building cash registers, computer vision Stuff for ATM machines and things of that sort. They were building some of the stuff that people would have in ambulances, like from Medtronic and things of that sort. So some medical devices. Yeah, they were building some advanced stuff there. 00:17:45 - Scot Wingo Very cool. Yeah. So you're down there and then what? I know you're at IBM for like a fairly about a decade plus. So what are some of the highlights from that? 00:17:55 - Igor Jablokov Well, I mean, halfway through my career there, I kept sending notes to an early AI team and they said, hey, if we let you lead that group, would you stop sending us bloody emails? So I ended up taking over that group in 2002 onwards. And that's where, you know, we essentially discovered the baby version of Watson is what we were working on. And that's, that's where I got the AI bug. But you know, AI is essentially multidisciplinary, so it has a little bit piece of everything. 00:18:29 - Scot Wingo Yeah. Put Watson in terms of where we are now in AI. So today's AI and it's such a weird term, right, because it's so encompassing everything. Yeah. Now when you say AI to the current folks, they think LLMs and ChatGPT and all that kind of stuff. But like that AI was more explain where Watson fits in this kind of hierarchy of AI. 00:18:48 - Igor Jablokov Well, back then it wasn't even called Watson, just like, you know, there was early experiments and other things. What I would say is that was more statistical methods. 00:18:58 - Scot Wingo And prior to that deep learning. Is that the same? 00:19:00 - Igor Jablokov No, no, deep learning came, came later. So whereas now you got your neural networks and transformers, you know, so that was a big difference. You know, Right now a lot of this stuff is transformer based. Back then it was still essentially feature engineering where you were trying to mimic the way that a human ear worked and building language models out of bigrams, trigrams, quad grams. Prior to that they were doing grammar based recognition. Sort of like when you would call a bank or airline and they ask you for dates and locations and things of that sort. It was finite domain, you know, is how it started. 00:19:43 - Scot Wingo Yeah. 00:19:43 - Igor Jablokov And then it started moving towards transcription or dictation based. That was more statistical. And then now you have all the transformer based stuff. 00:19:52 - Scot Wingo Yeah, cool. So you're at IBM, you're working on this, this, you know, so you started thinking, wow, it'd be interesting to talk to dolphins. And now here you are thinking through like natural language. Natural language ends up for lay people. It's a very, very hard problem. Right, so is that like, did that really grip you as part of this AI thing is that. 00:20:12 - Igor Jablokov Oh, yeah, because the thread that kind of got you. Yeah, the OGs and AI, there were three reasons why they were attracted to the field. And remember, there was no fame and fortune in this, you know, a quarter century ago. But there was three reasons why we were attracted to the field. The first reason was for accessibility reasons. You know, our chief scientist at the time that was attached to our group was a blind fellow. You know, he's now at Google. So accessibility is a pretty nice thing to do for folks, right, to open up the aperture and allow it for, you know, for children, for senior citizens and things of that sort. So that, you know, was, was enthralling. The second reason to work, work on it was for safety reasons. You know, think about, you know, people crashing their cars, texting while driving and things into trees and utility poles and things of that sort. So it creates safer experiences, especially as we were working with automotive partners. And then the third reason was to bridge cultural divides, you know, with machine translation. So that, that's what really attracted a lot of folks to the field way back when. And that's when we were discovering, you know, rudimentary versions of, of early AI assistance. And we had a lot of experiments. You know, we were hiding these things in Miami Children's Hospital, you know, to, you know, to put on, you know, LCD screens in there to start and stop surgical timers. We were hiding it on Wake Forest University's campus for things like, hey, where's the shuttle bus? You know, is the washer and dryer free? And things of that sort because coeds are obviously too lazy to go down the hallway and check for themselves. And the third place we hit, it was an Epcot center in Disney World. Part of the reason is to get it in front of children because they're imagineers. You never know how they're going to adopt technology and what use cases they're going to discover. And then in hindsight, I realized accidentally we were collecting information about impacted speakers. Essentially the amateurs that go left when they enter Epcot center and immediately go to Mexico and get wasted, rather than starting off easy, going to the right side and starting with Canada. 00:22:31 - Scot Wingo It sounds like you've, you've got some experience here. 00:22:33 - Igor Jablokov Yeah, I go, I'm not a big drinker, so I go right. I don't go left. 00:22:38 - Scot Wingo Yeah. But, yeah, trying to do the multicultural tour. They're doing the, the beer brawl. 00:22:42 - Igor Jablokov I'm just. Well, I'm just looking for maple syrup because I went to secondary school in Montreal, so that's that's, that's the siren call is Canadian sweets for me. Yeah. 00:22:55 - Scot Wingo Cool. So you're at IBM doing IBM things which is the opposite of being a startup founder. What kind of led you to the ultimately leave and start a company? 00:23:03 - Igor Jablokov Yeah, at the same time I, you know, it was around that time that I got my MBA at uncc. And then one of the things that was curious to me, we were running an internship program called Extreme Blue where we took a top MBA candidate and four undergraduate candidates and we created like little mini startups and we were hosting them at that time in Austin, Texas. And we took a lot of raw technologies that were mind blowing, like cell architecture, what people now know is the PlayStation 3, early AI engines and things of that sort. And we just handed it to these teams to see what they could do. Now a lot of these folks, you know, years later would become, you know, business leaders in their, you know, in their own right. Like for instance, the MBA candidate that we had, you know, ended up becoming CEO of the RISC V Foundation and now runs sovereign AI for Nvidia. But back then they don't even know what technologies they were getting access to. And when I saw what they did at the end of that summer, I'm like, holy smokes, these are five students. What if I took 5 of top people and put it together and got them the resources? I bet they can build something, you know, even snazzier than the snazzy things that these students have constructed as well. So that's where that was, you know, one side of the equation and the other side of the equation is IBM was decommitting from a lot of this early AI stuff because it was ahead of its time. The business case wasn't there to continue investing in it. But my spidey sense was going crazy like, oh, this stuff's going to be big. We just have to be patient and things of that sort. 00:24:51 - Scot Wingo Even pre Watson, they were starting to. 00:24:54 - Igor Jablokov Oh, they were waning. 00:24:55 - Scot Wingo De. Prioritize. 00:24:56 - Igor Jablokov Oh, they were prioritizing. They were moving people around, not investing in it and things because they had. 00:25:01 - Scot Wingo Was this in the Ginny Room. I forget. 00:25:04 - Igor Jablokov I think, I think it may have been like the end of Sam. 00:25:07 - Scot Wingo Okay. 00:25:08 - Igor Jablokov You know, days and things of that sort or early Ginny days. But, but yeah, they had a lot. 00:25:14 - Scot Wingo Of other things going on. They had sold off the PC, the division, the main problems. Yeah, they had like, yeah, they had a lot of issues. They were kind of fighting on multi fronts. 00:25:22 - Igor Jablokov Yep, yep. So I mean they, they had it all, you know, under. Under their auspices. Similar to Kodak. Right. Having a chance with digital technologies and HP or parc. Right. Having. Having access to wonderful stuff. 00:25:40 - Scot Wingo Yeah. 00:25:41 - Igor Jablokov That they just didn't know how to commercialize. 00:25:43 - Scot Wingo It's the Innovator's Dilemma. One of my favorite books. 00:25:46 - Igor Jablokov But in some ways, I mean, you know, dumb luck or providence, you know, My brother was actually getting his MBA at Wake Forest at the time, and he was going to submit a homework assignment for his entrepreneurship class. And I looked down at it and I'm like, what is this? He's like, oh, it's a titanium spinal implant. I'm like, you're not going to get that FDA approved for a decade here. Move aside. I was sitting in his kitchen and then I typed up multimodal portal. And he's like, what is that? I'm like, here, just submit that this is going to be a big deal one day. And so he submitted it as his homework assignment. And the professor said, holy smokes, this is a big idea. You should pitch it to our business plan competition. What's a business plan competition? Like, okay, we'll show up at Wake 4. So me and him showed up and I revealed it. The grand strategy for multimodal portal, hosted client server based. All this. All these words that don't make any sense to folks until I translate them to what you now know as cloud and AI assistance and things of that sort. We lost. But this professor ran after us and said, hey, you're onto something here. I do a lot of work with Nokia. You just have to simplify it for these folks to understand what this is. Pick one use case rather than all of these use cases. And I just shrugged it off and I'm like, fine, whatever, we'll do another one of these. So at UNCC in Charlotte, we did another business plan competition and we lost to a soybean. A genetically modified soybean is what we lost. 00:27:28 - Scot Wingo That's always humiliating to list us one time to a. They were doing software for administrative assistants. I was like, administrative assistants don't use software. They just use Google sheets. And they're like, I don't know. Yeah, it's a. It's a motivating. 00:27:42 - Igor Jablokov And did you lose? Was that the. The software that's in Greenville, South Carolina? 00:27:47 - Scot Wingo Yeah, I lost them. 00:27:48 - Igor Jablokov Yeah. 00:27:50 - Scot Wingo So I hope they did well. But it was like, it was my soybean. 00:27:53 - Igor Jablokov They. Yeah, they. Yeah, it was your soybean. I think they're one of the rise of the rest companies too, or something of that sort. 00:27:59 - Scot Wingo Okay. 00:27:59 - Igor Jablokov But yeah, yeah, that was your soybean. So anyway, I lost to the soybean. But get this. This guy came up to us afterwards and says, hey, I'm your first investor, first board member, and never publicly speak about this again. I'm like, who the heck are you? Apparently, this was one of the founding executives of Amazon.com and he just happened to be in Charlotte because he promised his spouse that they would get out of the rain for a spell while his kids were going to school. Most of her family was in Charlotte, and that was his penance for bringing him to Seattle for a while. And we ended up crossing paths, and that was a chain reaction that had me meeting attorneys and other investors and things of that sort, and we were off to the races. And that was in 2006. Now, what happened in 2007 was crazy because I'm going to reveal something and then I'm going to tell you why it happened. In 07, I presented at the very first TechCrunch Disrupt conference. And I walked out on stage. I pulled this razor flip phone out of my pocket, I spoke into it and it talked back. And Andreessen, Marissa Mayer Geek Kawasaki was there. Everybody was scratching their heads and they had no idea what they were seeing. I was not allowed to tell anybody that we were secretly working with Apple on Siri before the iPhone even came out. Now, how did we even cross paths with Apple? Well, that's because Steve Jobs hid the ipod R and D team in a nondescript building outside the Charlotte airport. They were here. I don't know that they're in the Carolinas. And so in our last startup, I had started recruiting people from Portal Player, Nvidia, Broadcom, Marvell, intel and things like that. So we had a lot more a subject matter expertise on the hardware side, you know, paired with folks coming from software side, from Nuance, from IBM, from Microsoft and things of that sort. So it was essentially a special environment. 00:30:10 - Scot Wingo The name of the company was Yap Yap. 00:30:13 - Igor Jablokov Yeah, we hadn't, you know, in hindsight, we probably could have come up with a better name or we're just doing trademark searches and things like that. What do you call a speech recognition company? And then, and then my brother's dog barked. I'm like, fine, we'll just call it Yap, you know, and that's. That's where it came from. 00:30:32 - Scot Wingo Sometimes people way overthink the name of their companies. 00:30:35 - Igor Jablokov Well, yeah, I mean, it was a ludicrous company in hindsight, considering how prominent it got. So, you know, When Amazon ended up buying our company five years after that TechCrunch conference, everybody thought we were an aqua hire. But guess what? People didn't know. We had almost 50 million users on that platform. Dozens of enterprise and carrier customers, including Sprint and Microsoft. We were the first to do streaming data before Apache Kafka existed. We were the first end to end system. We were using a type of neural network. Before Hinton's paper came out, it was the first shot fired on AI acquisitions by big Tech. And even though Cade Metz wrote that book about Hinton's company getting acquired in 2012, we were the first one to essentially go out. It was the first AI cloud company, the first hybrid. There were a lot of firsts. And that thing that we didn't know, we were just doing our jobs. 00:31:37 - Scot Wingo Yeah, how'd you have 50 million users? You were just kind of powering things like a series kind of thing behind the scenes. 00:31:43 - Igor Jablokov Not even AI assistance. It was actually coal mining. Yext was one of our clients. Right. Jesse Lipson was on their board. And a lot of voicemail attacks. Now I'm going to reveal something crazy because all of us as founders, as business executives in these startups, what do we do? We build products. Why do we build them? To have great experiences where we have end users that adore the product and then as an accidental byproduct, we get paid in order to reinvest and build the next version of this thing. That's not how carriers think. They couldn't care less that this was a better user experience. I'll tell you why this started selling to them like hotcakes. Once they derived metrics that an end user, essentially a phone line that had voicemail attacks provisioned on it was 50% less likely to churn away. They started buying it hand over fist. That's it. They bought us like an insurance policy. 00:32:51 - Scot Wingo To drive engagement of the other sides. 00:32:53 - Igor Jablokov Well, to keep people on their platform. Because if you think about a metric that they have to report up to their board is churn and anything that you can do to drop churn down by 50%. Oh, it didn't matter if it was a game of Tetris bejeweled voicemail attacks, you know, some sort of ringtone service and things like that, it was going to be radioactive and they were going to pick it up. Yeah, that's why it was starting to sell like crazy. Cool. 00:33:19 - Scot Wingo Awesome. Why did walk us through the acquisition of that. So how much did you guys raise and. Yep, let's start there. 00:33:25 - Igor Jablokov We raised 12 million, I think about 10 million in equity, 2 million in venture debt. 00:33:30 - Scot Wingo Okay. 00:33:31 - Igor Jablokov So it was pretty capital, efficient organization. But, you know, back then, ice cream cones, you know, cost 5 cents and. And things of that sort. It's not the crazy, you know, current days of. Of AI. 00:33:42 - Scot Wingo Was your brother part of the company? 00:33:44 - Igor Jablokov Yeah, okay. Yeah, yeah, he was part of it as. As a coo. And there were a lot of Mormons that work there as well. So we kind of jokingly refer to it as a Greco Mormon company. In. 00:33:56 - Scot Wingo In some ways, one side loved coffee and crazy stuff, and the other side was like, you know, sticking. Sticking to other stuff. 00:34:02 - Igor Jablokov Yeah, well, I mean, I was practically teetotaler, you know, that whole time and stuff like that. That's. That's the culture that ends up building there. But, yeah, you know, everybody was a hard worker there and we were really excited and we had no idea that we were at the dawn of such a big deal. We were just, you know, you know, I always. I always tell people, you know, founders don't need a compass. You know, it's just we just go out there and we just survive on Spidey sense and intuition and stuff like that as well. That's where we're good at. It's like, it doesn't bother me to have two engines out, bird strike a la Captain Sully. It's like, I'll find you a place to land. You may not like the landing and it'll be in water or in gravel or grass or a dirt road, but I promise you, you'll end up landing. Always kind of trust, folks, intuitive sense, because I can't find data for things that don't exist yet. You know, it's like, I'll be making it up. 00:35:02 - Scot Wingo Yeah. This is what drives me crazy of. I'm sure you've had this before where VCs will come in and they'll want to do this big TAM analysis of. You know, imagine the TAM analysis of yap of day one. Yeah, it's like zero. Right. Like, there is no addressable market. Like build. The whole point is we're building addressable market. That always drives me crazy. 00:35:17 - Igor Jablokov Now it's like 8 billion people can have access to it. Right. And it's in every Android device, you know, every iPhone device, every social media app has it and things like that. It's just essentially commoditized. But back then. So my family members kind of hate when I do this to Uber drivers. I'm sitting in the back and they'll do a voicemail or Voice to text, right as they're driving and stuff like that, responding their friends or family members. And I turn to the driver and I'm like, the person who invented that must be a genius. What do you think? Do you ever wonder about this person? And of course my family members are in the backseat, they're rolling their eyes. I'm like, I wonder what that person is doing right now, you know, Must. 00:36:00 - Scot Wingo Be a super nerd. 00:36:01 - Igor Jablokov Must be a super nerd. I wonder who this person is. And of course they're just rolling their eyes because back then I heard every negative thing you can think of. You know, nobody's going to talk to a phone like that. You know, how are you going to get it distributed? You know, there were, you know, there wasn't marketplaces, how do you distribute this? There was no business model for IT devices you had to write in brew for the Verizon CDMA style devices you had to write in JTME Java based code for that. And then there were JSRs. Not all of the Java phones allowed you to record audio and things of that sort. So it was messy. It was pretty messy back then. 00:36:45 - Scot Wingo That's the solving messy problems. If it was easy, other people do it like you gotta solve the hard stuff. 00:36:50 - Igor Jablokov I gotta tell you, it was accidental that we turned into the first ever AI cloud. Completely accidental. So let me tell you what ended up happening. I kept failing at VC pitches. So I'd show up at General Catalyst. I remember showing up in Seattle and I'm like, hey, you know, I'm working on this thing called a multimodal portal. Again, we don't know to call it an AI assistant. Siri doesn't exist yet. None of this stuff exists. Nobody knows what this stuff is. 00:37:15 - Scot Wingo When you say multimodal, you mean like voice, text, everything? You didn't mean video, but like you probably had a vision. Video would come. 00:37:21 - Igor Jablokov Yeah, yeah, that's like anything input and anything output. Yeah, multi. Multimodal portal. And the portal meant think about like old school Yahoo Portal. Like you could get news, weather, sports, music and things of that sort. It could be anything, anything input, anything output to get anything information. Right. Doing messaging as well. And, and they're like, I don't get it. You know, I don't get what this is how you can distribute it. What's the business model? There's no support, you know, for recording audio and things of that sort. But I have this voicemail company. Can they send you the voicemails and then you just send back the textual representation and I kind of shrugged. I'm like, yeah, fine, they can do that. And then I would go to the other place and they're like, we don't get it. I don't understand what this multimodal portal nonsense is all about. Again, didn't know to call it AI Assistant. They're like, but we have this call mining company. Can they actually just send you the phone calls and you can just hand back, you know, the textual representation of the phone calls? I'm like, sure, fine. Then I went to the third vc. They're like, hey, I have this search company. Can they do voice searches where, you know, people can talk and. And you can do that? I'm like, fine. And then I went to the fourth one. They're like, well, I have this messaging, you know, company, you know, similar to WhatsApp and stuff like that. Can people just record their messages and stuff like that? I'm like, yeah, fine. And so we accidentally became this AI cloud company where people would connect to us via APIs, all on the journey of trying to get the AI assistant funding and then failing. It ended up being the biggest customer, though. Oh, yeah. Oh, yeah. We kept winning over and over again. But this is why whenever I meet founders that are wary of fundraising and stuff like that, I'm like, what are you talking about? You're going to meet 100 different entities now and you're going to accidentally discover clients and partners in their portfolios. This is a tremendous opportunity. Every two years that we end up fundraising, you end up having your eyes open and getting stress tested and things of that sort. You got to embrace the suck and find the utility in it, rather than just finding pain in it. 00:39:39 - Scot Wingo Yeah, yeah. Every nose a pathway. Yes, yeah, yeah. Okay, so then you're deep into yap. How did Amazon come about? And walk us through kind of like your choice to sell. And I don't know if the result of that has ever been made public, but whatever you can tell us around, like, the outcome of that. 00:39:58 - Igor Jablokov So I didn't want to sell the company. By then, I was already minority shareholder through all those rounds of investments. But for the investors that we had, think about the time frame. Most of their portfolio companies didn't work out because of the 2008 downturn. And so this, you know, here we are getting offers from. Nuance wants to buy us. Google wants to buy us. 00:40:26 - Scot Wingo Around that time, Nuance made this thing called Dragon. 00:40:30 - Igor Jablokov Dragon. Right, right. So Nuance wanted to buy us. Microsoft wanted to buy us. They lowballed us. Google wanted to buy Us. 00:40:39 - Scot Wingo Is that Ballmer era, or is Sacha there there at that point? 00:40:43 - Igor Jablokov No, Sacha wasn't there yet. Yeah, it was. It was pre Sachet and Google wanted to buy us, so they called us up to their east coast headquarters up in New York, and they push a paper offer across the desk. And they're like, you have to sign it on the spot. I'm like, what are you talking about? I gotta take it to my board. They're like, no, you gotta sign this piece of paper on the spot, Larry. You know, we talked to Larry. 00:41:12 - Igor Jablokov We want to acquire you. I'm like, I can't do that. They're like, no, we're going to go to paper. Because I guess they were smarting from Yelp, leaving him at the table, so they were worried about leaks and things of that sort. And so I waited until there was only one attorney left in the room. And I kept feeding my brother water because I know he has a small bladder. 00:41:33 - Igor Jablokov And then eventually I said, hey, this is such a fancy headquarters and stuff like that. I think we're going to get lost. Can you please show them to the nearest restroom? And the attorney happily obliged and took him over to the restroom. And I just looked at the pieces of paper and just started photographing them and then sending them to my board in terms of, hey, decide, you know, whether we sign this or not as well. 00:41:55 - Igor Jablokov I'm like, that is weird. You can't not share, you know, you know, a financial instrument like that with a board in order to, you know, make a decision. But it was, how did we end up meeting Amazon? Here's why. Remember when I told you Steve Jobs had the ipod R and D team? So I asked, you know, our head of engineering who came from that experience, he's one of the Froh fathers of the ipod. 00:42:22 - Igor Jablokov It was two folks from Portal Player, two folks from Apple that did it. And I said, how did Apple discover you? He said, well, we were working on tabletop radios, and we were in the back of ces. We just had a lone table where he had a tabletop radio. And these Apple folks crossed paths with us and said, hey, what is that? They're like, oh, it's a tabletop radio. Like, no, you're not going to be doing that anymore. And it was just completely dumb luck and providence. 00:42:53 - Igor Jablokov And so when I heard that story, there was one week in the early part of 2011 that I remember there was three different conferences that week, and everybody wanted to go to the big one in Vegas, and there was a smaller one I think in New York and the smallest was in Florida. And I had to kind of be a Hebrew king and kind of decide for us. I'm not going to. I personally, as founder, I'm not going to go to the big one. We're going to go to whichever ones we're geographically closest to. So Felix, you're going to go to the Las Vegas one because that's closest to you. From San Francisco, I'll go to New York. And we're going to send a third person down, Victor, we're going to send him down to Florida. 00:43:42 - Igor Jablokov Sure enough, at the smallest, humblest conference, that's when Amazon crosses paths and says, hey, what are you guys working on now? I also have a suspicion that they were monitoring our AWS usage because we had all physical servers, but we figured out how to do hybrid cloud where during the peak busy hour, twice a day, instead of buying metal for rush hour in the beginning of the day and end of the day, it would just blurp into aws. And maybe they also said, hey, this is weird. What are these two hours a day where a lot of use ends up flowing through our virtual machines and things of that sort as well. 00:44:25 - Igor Jablokov So I have a suspicion about that. That's not confirmed. But yeah, that's how we met them. It's just by, you know, as Gary Schindling was once asked by an up and coming comedian, what's a shortcut to comedy? And he's like, there isn't one. You just have to be out there. 00:44:43 - Scot Wingo Yeah. So Amazon approaches you, why'd you end up selling to them? Was it just. Did you run a process and have some bids and the board just kind of like went with the biggest one or there was some kind of an alignment thing. 00:44:53 - Igor Jablokov Yeah, I mean there was some alignment, but like I said, we had contacts at Microsoft, at Nuance, at Google that we knew were interested in us. And so I kind of ran that, ran that play myself. We had no bankers because I wanted to figure this out, you know, for ourselves. But then, then we ended up getting acquired. Like I said, most of our investors portfolios, you know, weren't doing all that well and, and they were euphoric. 00:45:21 - Scot Wingo There's a little pressure to it. Sure would be nice if we could get it. 00:45:25 - Igor Jablokov Oh, in height. Yeah. In hindsight we would have become like Twilio had we kept going with fairly massive enterprise value because we're ahead of it. It's time. Or similar to bandwidth. Right. And their prominent position in, in the, in the Carolina ecosystem. But you know, that's, that's what happens when, when people get impatient. 00:45:45 - Scot Wingo Yeah. All that was in Charlotte. 00:45:47 - Igor Jablokov All that was. It was headquartered in Charlotte. Most of our R D team was in Boston area in sales and marketing was in the Bay Area. It's my fault that everybody in Massachusetts had to start paying sales tax because the minute that they acquired our team in Boston, that ended up meaning that Amazon had a nexus there and everybody there, you know, that for those holidays had to start paying taxes on their, on their purchases there. 00:46:16 - Scot Wingo Cool. Well, I don't think a lot of them listen to the pod, so I think you're safe. 00:46:18 - Igor Jablokov Yeah. Yeah. 00:46:19 - Scot Wingo Was this a good exit for the investors? So you have 10 million investment. Have, you know, can you say how much it was sold for or. That's. 00:46:27 - Igor Jablokov I'll say this. Some, some of them made a 10x+ thing. In fact, I was, I was at a conference here recently and this, you know, fella came up to me and said, hey, I, I heard that you were giving a keynote here. And so I went back in my, in my records. And that one thing was the most return that I've ever, that I've ever gotten in, in my career. And you used to run the pension fund for RJR Nabisco. So you put two and two together. It wasn't, it was all right. It was all right. So we did right by folks as well. But look, it looked hairy down to the end. I mean, we only had $14,000 left in our bank account on the day of close. Had it actually clipped 24 hours later, we would have had to trigger payroll and the company would have been bankrupt. So that's why I tell people. Let's, you know, I don't, I don't, I don't know what founders are, are made of, you know, not sugar and spice and, and nice things, but we just don't panic, you know, just. Yeah, as you know, I swatter for veins, you know, in, in your arteries and veins and stuff like that. It's just, you're just creating logical solutions to problems as they come up. And you know that problems are always coming up. Is this a new day? New problems, and you're just solving them? It's, you know, you don't, you don't get emotionally damaged that, that these things are out to get you or anything of that sort. It's just another day at the farm, basically. 00:47:57 - Scot Wingo Yep. Yeah, just a little piece of flotsam and jetsam trying to navigate the waves. 00:48:01 - Igor Jablokov Yeah, the. 00:48:03 - Scot Wingo Okay, so Amazon acquires you. How long did you stay on and work for Amazon. 00:48:07 - Igor Jablokov I helped write the strategic plan for it and then immediately pivoted and at that time got an Eisenhower fellowship that was chaired by General Colin Powell. So I wanted to do some. I was worried that I would get typecast in AI and I'm like, oh, gosh, you know, I'm going to turn into like Jim Carrey and the Pet Detective, where people are going to think I'm this AI person the rest of my life. And so I did a hard pivot towards national security stuff, you know, trying to understand these entrepreneurial ecosystems in the Middle east and things of that sort. Little would I know that years later, those two worlds of the public sector and AI would collide again. So that detour that I thought was a detour was sometimes the straightest line between two points is a curve and they end up colliding anyway. 00:49:03 - Scot Wingo And you are Jim Carrey, the Pet detective. 00:49:04 - Igor Jablokov Yeah. And then. Well, that's. That's what ended up happening when, when this company was found founded is I literally dreamt it up. I literally dreamt it up one day. 00:49:16 - Scot Wingo I definitely want to catch that. But, like, I don't think people. I know this, but what did Amazon do with Yap? 00:49:22 - Igor Jablokov Oh, that's right. 00:49:24 - Scot Wingo Let's just get that on the recording. 00:49:25 - Igor Jablokov Yeah. So what Amazon ended up doing with Yap, that any. Nobody had any idea was Alexa. So Alexa is my older sister's name, which is a coincidence. And the code name for it was Prion, which we ended up reusing for the next company, I think the Amazon Echo and Alexa didn't come out until 2014. And so for about three, four years after that, everybody was puzzled. So sometimes you read the news in there and they're like, why did they buy them? What's going on? And they try to keep it on the down low as long as possible. And so there was a lot of mystery associated with our company. In fact, TechCrunch wrote an article that we entered the deadpool thinking that we went out of business because we weren't allowed to tell anybody that. That we were acquired. 00:50:15 - Scot Wingo Cool. Okay, so you basically wrote the technology that powers. 00:50:19 - Igor Jablokov Yep. 00:50:20 - Scot Wingo Alexa. Okay, so you're dreaming. It's 2016. 00:50:24 - Igor Jablokov Yep. 00:50:25 - Scot Wingo Time frame? 00:50:26 - Igor Jablokov 2017. Time frame. 00:50:28 - Scot Wingo Yep. 00:50:28 - Igor Jablokov And I get this peculiar dream of a new. A new AI architecture. And I'm like, that was pretty weird. So I ended up calling. 00:50:39 - Scot Wingo Were you passive in the stream? You just like, saw it on a whiteboard? 00:50:41 - Igor Jablokov I know, I was, you're in the. 00:50:43 - Scot Wingo Future talking about it. Like, did you project yourself forward. 00:50:45 - Igor Jablokov I was sitting in my office and I had an Amazon echo. There was Siri there. There was. I think Cortana was there, and Watson was there. And I was talking to all of them, literally simultaneously talking to them. And they were all jibber jabbering back. And I'm like, wait a minute. All of you shouldn't be talking to me at the same time. And then that's where I saw a new architecture for essentially fusing all of these disparate things in together. And so the following day, I'm like, that was weird. So I sketched it down on my iPad and I picked up my phone and I called buddies of mine that were Alexa, and I'm like, hey, I think this is going to save Alexa and adapt it towards this next version. You should steal this idea. And they said, holy smokes, we'll resign and join you if you get this. 00:51:38 - Scot Wingo Funnel turned into a recruiting call. 00:51:41 - Igor Jablokov Yeah. And I'm like, I don't want to start another company and deal with investors and stuff like that. Are you kidding me? And I hung up the phone. I'm like, that's crazy talk. 00:51:50 - Scot Wingo I think implicit in there is they probably knew Amazon wouldn't do anything with. With it. Right? Amazon's become kind of a very big company. They. They're not going to just take some random Eagle and run with it. 00:51:59 - Igor Jablokov Yeah, but the point is, it's like, you know, it's like. It's like that book. Steal this book. Yeah, it's like, steal this AI. You know, I was trying to give it away for nothing. And then I called a second buddy of mine that was on the, you know, the Watson Jeopardy. Team. And I'm like, hey, I think this will save Watson. He's like, hold on, let me send you over to the cto. And I told him what the idea was. I'm like, you guys should steal it. And he's like, I'm in. I'm like, what do you mean you're in? He's like, yeah, this is big. This is going to be the swan song to my career. And I'm like, it's like, no, Facebook's trying to recruit you to run their AI. Microsoft has been trying to recruit you. Google has been. There's nothing to be in on. I don't want to do this. I don't want to deal with investors and things of that sort. This is crazy talk. And then. And then I called a third buddy of mine who was reporting to the Cisco CEO, and he's like, we'll buy it, we'll buy it for this amount. And I'm like, oh, now you screwed me. I know it's worth 10 times more than you. Just then you just offered to buy it. And so now I'm forced to do this. And so that was in the early part of, of of 2017. And then, and then I started you know, obviously traveling to Boston and New York and, and the west coast and things like that to kind of socialize it amongst, you know, diff different investor contacts that I had and, and started doing the March of Dimes on, on you know, getting an angel round done and, and starting to recruit our first folks, you know, starting in Seattle and then eventually down here in New York, you know, Bay Area and the like. 00:53:39 - Scot Wingo So I remember on that timeframe it's maybe been before Prion you were talking about it's always fun to have coffees with you, so I have to have a venti kind of latte to just kind of like be able to hang with you. And then you know, so you were talking about, you know, we're going to have a future where you know, every, they'll just understand us and we'll have conversations and you want to talk to your. Maybe it's a device, who knows what it is but you want to say hey, what were sales yesterday and what's trending here? Yeah, you know, I think the vision back then that became Prion was like almost an assistant you would talk to about business stuff. But that. 00:54:10 - Igor Jablokov Correct. 00:54:11 - Scot Wingo But what's interesting is embedded in there is what today we would call LLMs which have come to fruition. But you were talking about this in like 2014. You and Robbie were both like on that pretty early. But then like, you know what we'd call today? Retrieval Augmented generation. So then you were going to take this, you know, kind of the consumer piece, apply it to a business, run it through the businesses data and then come out with, you know, something on the other side. So here we are 2025 recording this and kind of 10 years later things have caught up with you. 00:54:41 - Igor Jablokov Yeah, and remember we always wanted to be anchored to facts. Yeah, right. Because people don't trust technology. People trust other people. And so I always wanted to see the authorship of the underlying assets that we're going to be giving them. You know, business solution. You know, if you're going to give, you know, you're negotiating a contract, giving somebody, you know, a medical diagnosis or things of that sort, you better know why. You know, the, the machine is giving you, you know particular answer. So it can't be hallucinatory nonsense and, and things of that sort. 00:55:17 - Scot Wingo Yeah, that's where the, the database grounds it on. Yeah. So. So even in today the LLMs do all these crazy hallucinations, but the database tends to ground it. It still can have hallucinations, right? 00:55:26 - Igor Jablokov Yeah, yeah. And, and, and database we're using in an abstract form, of course, it's a system underlying systems record, SAP ServiceNow, Salesforce, you know, things like PowerPoint, PDF, Word File, Web pages, audio, video, image, text stored in sharepoints, box folders, Documentum, Google Drives and things of that sort. There's a lot of unstructured semi structured and structured content in, in all of these different, different repositories and what we foresaw now in the early days, you know, the thing that I thought I would be doing was semantic integration with the AIs and having them all work together as a team, but knowing that one is going to become an expert in one thing, the other one's going to become an expert in another thing. 00:56:10 - Scot Wingo Kind of like an authentic model is another thing that we would call that today. But you were calling, you were just thinking of it in a different framework. 00:56:17 - Igor Jablokov It was exactly the same thing. I always joke with people, you know, don't. If you show me Lang Chain, you know, all I hear is, is, you know, Shania Twain songs in my head starts playing, start playing. Or it doesn't impress me all that much because that was literally in our first presentation materials from March of 2017. 00:56:37 - Scot Wingo LangChain is a framework for building agent based systems, correct? 00:56:41 - Igor Jablokov Yeah, yeah. So we kind of foresaw. 00:56:43 - Scot Wingo So basically you had the vision of the architecture when you had that dream. You're basically envisioning an architecture around AI agents working together in some kind of a collaborate. Not a collaboration, but a orchestration layer. 00:56:56 - Igor Jablokov Oh and we even had a prototype of it. It was leading to higher accuracy, you know, by orchestrating all of these things together. I remember the Gray Croft folks in New York area ended up seeing it and that's why they were the largest investor in our, in our seed round that I just met through happenstance while I was in Atlanta one time and visited with them in New York. So I give them credit as part of that seed round. And you have to inherently give people who do invest in earlier stage companies like Tweener or otherwise credit because founders don't know what to call things because there is no words yet. The media hasn't figured out to call something cloud yet or an AI assistant or an agentic or LLM. We had language models in our last company. We didn't publicly talk about language models, acoustic models, lexicons, bigrams, trigrams, quad grams and things of that sort. It was ahead of its time. They were all in there. Nobody knew what an end to end system was or neural networks were and things of that sort. You just didn't talk about it because that's like talking to. About carburetors, you know, to civilians and things of that sort. If you're an auto mechanic. 00:58:18 - Scot Wingo Yeah. This time around as an outsider when you started pitching, now people are listening to you. So you know, your first round, you did like a seed round that was relatively large for a seed round. And then you've raised a lot of capital at Prion. Walk us through. You know, happy to go wherever direction. I want to make sure we have time to talk about your vision of the future because you're 10 years ahead. I always like to kind of, you know, give people a little insight into that but like give us a quick summary of, of where Prion is now, how much you've raised and yeah what you're doing. 00:58:48 - Igor Jablokov So, so I did my seed round and you know, Carolina Angel Network is, is part of the seed round amongst amongst other folks. And Graycroft and Revolution Rise of the rest showed up. Two Sigma showed up amongst others. And then I get this phone call and I answer it and I'm like, who the heck are you? He's like, hey, I'm really excited about what you're working on because I work closely with Peter Thiel on Palantir and stuff like that. And I think you're onto something and I know that you're going to develop this differently than what's going to end up happening in other markets like Silicon Valley and things of that sort. And he's like, I'll help you run the company, join your board and I'll introduce you to more investors. And that's when Jim Breyer ended up joining. Now Jim, for those of you that don't know, he was on the Harvard board, he's on the Stanford board, he was the first institutional check in Facebook and served on the Facebook board as well. Fantastic guy. And the person that called me that was J.D. vance and just was his, his firm. 01:00:01 - Scot Wingo Was he at the firm he had started or he was working with? 01:00:03 - Igor Jablokov No, that was the prior. He was working with Steve Case of AOL fame. And then you know, then after that time, you know, folks like Eric Schmidt invested and Sam Palmisano, the ex IBM CEO and folks of that ilk. 01:00:19 - Scot Wingo It's a nice callback. 01:00:21 - Igor Jablokov Yeah, yeah. I mean in hindsight. And then, and then we got introduced to this new dual use fund called USIT that was in summertime of 2022. It's not quite constructed yet. So they end up investing in a convertible instrument as they're building up their, their fund. And then in 2023 they ended up leading our B round, which was a hundred million dollar round. We'll probably end up doing our C round this year sometimes is what we're now focused on. So you know, that's, that's the funding that we've had still relatively capital efficient for one of these newfangled AI companies. 01:01:05 - Scot Wingo Yeah, hiring, hiring people that can kind of grok what you're building is they're few, they're extremely scarce and thus very expensive. 01:01:12 - Igor Jablokov Well, not only that, I mean, I can't grok what I'm thinking either, so. So it's like on a panium and in some ways, you know, just have to stare at ceilings and stuff like that. And like how does all this stuff fit together? But look, when, when my grandparents, you know, came to the U.S. after World War II, you know, my grandfather, you know, was a watchmaker then, you know, that's what he did for a living. And, and for me, I kind of see these, these things, as you said before, as having a system view of it as well. I think I relish the complexity of seeing how this gear connects to that gear, connects to this gear, connects to that gear, connects to this gear. It connects it. The more pieces the better because that creates a technical moat. And it doesn't, it doesn't stress me out that it, that you know, to service a client, a partner, you know, you know, to get, you know, certain types of certifications or accreditations that we need in order to get everything that's difficult. Instead of whining about it, about how long it takes to actually put that piece in there. You got to think like the third little piggy that's building with brick that yeah, you're not out there partying and it takes you longer to actually construct these things together. But boy, is your organization far more resilient on the other side for having those governance structures in place and technology that people just can't download from open source, you know, assets and be up and running. Now is that harder nowadays because a lot more people are paying attention to the field? Absolutely. And that's why, you know, when you talk about, hey, what's your vision of the future. In some ways, I'm talking less about what it is now, just from the standpoint of, you know, I'm not wearing a tweet coat and being a college professor, trying to train everybody up on it as well. We're still a commercial entity, so we're keeping a lot of this stuff for ourselves in terms of where this stuff is going to end up going. But I think we see some light at the end of the tunnel that's going to rapidly increase the enterprise value of the company here in short order. So that's what we're working towards. 01:03:25 - Scot Wingo Cool. And then I was at the bank of America conference where you spoke and you were talking about, I think you guys do some stuff and you hinted at this, you do stuff in the public sector. That's an example of, you know, a lot of startups, VCs would tell you, oh, don't mess with the government, you know, because there's like, there's complicated procurement, there's their own cloud you have to get certified for and all these things. But to your point, like, if you can get through all that to the other side, then, you know, you're, you're, you've gone through this. There's a bit of a moat there because a lot of companies can't make it through to that stuff. 01:03:56 - Igor Jablokov The one thing that, again, this is why, you know, for those of us that have done more than one of these, I remember being on Sand Hill Road even, even with that last startup, and I would visit very smart individuals, you know, in, in each of these firms, the Sequoias, the Kleiners, the Mayfields and things like that. And one of the things that I realized in hindsight after visiting with all of them is they all just told me contradictions, you know, one told me to do more enterprise, another one told me to do more carrier, another one told me to do more consumer, another one told me to do more developer led. And what I had to do then, you know, because, you know, I had to give them, I mean, I had to listen to what they were, they were saying because obviously it's coming from an experience vantage point. So what you end up having to do is run a portfolio yourself and say, we're going to do a little bit for developers, a little bit for the telcos, a little bit for enterprises, a little bit of our own, you know, direct to consumer app as well, and then just run these experiments and then double down on your major miner once you see which one of these four paths ends up doing something Prominently for you. Same thing here. I mean there's every single one of those paths. A direct to consumer app, something for enterprises, something for partners, something for developers and things of that sort. You run those plays and then essentially reinforce the ones where you start seeing things. But I don't have a crystal ball other than knowing that those are at least the four patterns, major patterns that exist. 01:05:27 - Scot Wingo Yeah, the weird thing about founders, and let's see if this resonates, we live in this bubble of risk all the time, but we're really good at de risking, right? Oddly, that's the oxymoron other thing. So in a way you de risk it by having a little portfolio of things. Right. So you're building prion, you're in maybe a government vertical, maybe a finance vertical, seems like that'd be an obvious one. So on and so forth and then you run experiments. But you know, just like managing a portfolio, some are going to win and lose, but you kind of have to have. If they all lose, you're hosed. But it's better than putting all your eggs in one basket. So as you get to a little bit later stage out of product market fit, you can start to do some things to mitigate your risk. 01:06:05 - Igor Jablokov I don't worry about risk because, I mean, heck, I've been on Ukrainian front lines before. I will say this, I think that, I think you started, you know, stating this metaphor before, which is Bezos once said, you know, in a large, you know, institution you may hear 99 yeses, but you hear one no and, and your project or program gets nixed. That was my experience at IBM. Yeah, in startup land we hear 99 no's and 1 yes and, and we get our funding and we get to be part of the great, you know, evolutionary drive of, of innovation and the stuff that gets built down here. I always tell people there's, there's, it's a double edged sword. You know, the west coast, you know, you have to think of animals there evolving in a rainforest. They have plenty of food and water. You know, in the southeast, it's a desert, we just don't get as much capital, especially in the earlier state stages that they can get. But as a result of that, it actually forces certain adaptations that years later end up being more valuable. For instance, because we didn't have all this excess capital for compute, we had to derive more energy efficient algorithms as a result, energy efficient algorithms allow you greater scale and it also means higher security, like that wired guard protocol, you know, is actually better than open VPN and things of that sort, but it has less lines of code. So our stuff ends up being more secure, higher scale. And it was a forced adaptation because, you know, we didn't have as much capital as the West Coasters. 01:07:48 - Scot Wingo Yeah, our startups end up being scrappier too. 01:07:50 - Igor Jablokov Yeah. 01:07:50 - Scot Wingo Like in the founders end up being, you know, they can handle ups and downs more because they kind of have to. They don't have the luxury of sitting on A, you know, $800 million, you know, you know, cushion of funding. So, so, you know, Prion's awesome, but I want to, you're, you're also a leader in, in kind of thinking around where AI goes. Do you believe? You know, so, so Prion aside, obviously, so this is just Igor kind of pontificating. Do you, do you buy into this, you know, AGI, ASI and some of these kinds of things? Where, where do you see this, like, you know, 10 years maybe too far with the rate of change we're experiencing, like maybe like five years from now, are we going to be, you know, and then what do you think that does? There's a lot of people. There's, there's a deceleration people and acceleration people. This is the Silicon Valley thing. But at the core of that argument is we should stop right now. The deceleration people say this is scary. Let's stop. Because like, you think through AI and if we get to this, some kind of super intelligence thing, a. There's like always the Terminator thing, which is kind of, you know, pretty overdone. But, but then there's also like, the more practical is like what happens to jobs and white collar jobs and those things riff on that a little bit like what's, what's going on in Igor's head on these topics. 01:09:05 - Igor Jablokov I remember Robbie Allen invited me to Carolina basketball game once and, and the, and there were four of us, you know, standing, standing together and, and I just started randomly laughing and he's like, what's, what's so funny? I'm like, you know what if the Terminator would appear right now at this moment in time, I wouldn't know which one of the four of us it was after, you know, so usually he. 01:09:33 - Scot Wingo Broadcasts, which is good. 01:09:34 - Igor Jablokov Yeah. So I'm like, I just started laughing, finding the humor. Look, there's certain AI companies out there that run like criminal organizations. When you actually peel them apart, you know, they try to use highfalutin words and they can't admit that it's just a cash grab for them. Right. And it's a pump and dump scheme in some ways. And I can prove it by essentially relaying five different taboos that they breached. So if you would have asked me what surprised me about the last 24 or 36 months, it's the fact that so many of these taboos that were norms in the AI industry were breached. The first one is, I'm sorry, you can't run a nonprofit research institute and then all of a sudden become a for profit. Otherwise let's all start that way. Let's all create nonprofits, allow investors to contribute, and take the immediate tax write off. And then five years later, 10 years later, when we hit something big, we'll just change our status, bring you all on the cap table, and we're off to the races. It's a win win scenario for everybody, right? So that's the first taboo that was broken. Then the second taboo that was broken is of course, because now they had this excess GPU capacity because again, it was before, you know, the AI roller coaster really started. They're able to do what, crawl the whole world and it's like, hey, let's take all this copyrighted content and see what happens. We were never allowed to use copyrighted content. 01:11:06 - Scot Wingo Forgiveness in the Sentinel. 01:11:08 - Igor Jablokov Yeah. So anyway, they started ingesting the copyrighted content for films and books and images and things of that sort and just sweeping it all together because they could. They had this spare capacity. Then the third taboo that was breached is called the alignment problem. That's what you know as hallucinations. You know us as computer scientists, computer engineers, electrical engineers, and mathematicians. You know, we were paid on ones and zeros. You press a button, the lights go on. You press a button, the lights go off. It wasn't this maybe thing that kind of, maybe I'll tell you the truth, maybe I'll make stuff up and things of that sort. That's a third taboo that was broken, right? Is things that didn't act like the normal deterministic systems that people expected. The fourth taboo that was broken was RLHF, reinforcement learning with human feedback. This is fancy pants talk for the fact that it's not an encrypted connection like you interacting with a Google search that's just stored on these physical servers. It may be going overseas to contact centers in Kenya or other places, or your prompt and the data that you may have uploaded now gets spilled. And so a lot of attorney, you know, acquaintances and friends that I have are like, there's not a day that go by where we don't have to write a breach notice that client data or employee data, you know, ended up somewhere in one of these AI models. 01:12:32 - Scot Wingo Yeah. Let me boil that down. So you're working at a company and you upload some code or a document and into the little prompt window and you're. Your view as the user in that scenario, you think that you're. That's going to be kept private and not trained into the AI system. But you know, the. What they're. What, what some of these things do is there's this gray area where they're actually kind of training the system on them. 01:12:55 - Igor Jablokov Correct. And there may be human beings overseas actually seeing it. And then the fifth most despicable thing is the, the is what connects to your original question. Some people are then stating that they're working towards an AGI, which is nonsense. AGI is not going to exist with this type of underlying hardware. As a result, certain people start thinking it's divine. Then their scientists go on a lecture circuit and start talking about how they use it for mental health support. And then you have teenagers and folks that start committing suicide thinking that this thing is all seeing, all knowing and they start harming themselves because they don't have a good support network around themselves as well. And they wash their hands hiding behind, you know, 2:30 or things of that sort, thinking that, hey, you know, it's sort of like the social media companies washing their hands. Hey, we're not the ones that, you know, attack teenagers and children and things of that sort. And then, and then all bets are off. We haven't even figured out how to legislate around all of this stuff and regulate it. 01:14:03 - Scot Wingo Yeah. So the anthropic people are a set of folks that left and started their own thing solving this. Do you kind of put them in the same bucket or is there like a. You think they're all kind of in this mucky gray area in your. 01:14:14 - Igor Jablokov No, I mean it's certainly a spectrum, you know, certainly a spectrum. Nobody's all one way and nobody's all another way. You know, it's just, you know, degrees of separation. Obviously, you know, Apple would be an example that's, you know, stepping carefully into these technologies. But even they're getting in trouble with the new summaries and things of that sort that aren't quite perfect. And then, you know, people are always going to try to do bad things on generating images and things of that sort. And then they're kind of allowing Siri to spill into and leverage ChatGPT. So they're still essentially giving a window, you know, into those technologies as well. It's, it's, it's not clean, you know, for, for anybody in the, in the field right now. Which is why we've never been a B2C company. We've only been B2B. We knew that we were going to be the adulting of AI. We knew that people were going to use us in hospitals and airports and more serious pursuits like power plants and things of that sort. And so our thing was designed from the get go to be think of this way. Enterprise AI has to balance four things. Accuracy, scale, security and speed. That's it. Those are the four things that our chair sits on. And if you think about what's different between B2C and B2B, a buddy of mine that used to run AI at Twitter said, hey, what's the difference between consumer and enterprise AI? And I told them this. I'm like, you guys try to maximize the amount of time that people spend on our platform. On your platform, right? That's all you're focused on, maximizing the amount of time. For me, I try to minimize the amount of time a physician, attorney, a nurse, you know, a factor worker spends on my platform. I'm trying to get them, you know, the answer, perform a workflow as fast as possible and then let them get back, you know, to their, you know, core pursuits and things of that sort as well. So it's a completely different ball of wax in terms of, you know, how our team gets measured in terms of fill rate, completion and stuff like that. Whereas their job is to get somebody to spend a minute, two minutes, five minutes, 10 minutes, an hour, two hours, three hours, 10 hours, you know, you know, a day, two days, all week, you know, all month. Yeah, yeah. 01:16:33 - Scot Wingo They want their thing to be engaging and have a follow on question and whatnot. So, okay, so then it sounds like you're predicting the current class of people are going to kind of fall into some, some issues with what they've done. But then like that aside, is there going to be an AGI type thing or. No, you don't. You think that's kind of a bit of hogwash. It's like it's hard to define what it is, right? 01:16:55 - Igor Jablokov Yeah, yeah, it's, it's, it's, it look, it's, it's. What's the root of consciousness? I mean, they're even thinking humans may be quantum computers because they're tube of tubes where the creation or collapse of superposition, you know, is, is where we exist and that means technically it can be quantum entangled with other people and, and your pets and family members and business colleagues and plants and, and, and places that you've been in, things of that sort. We don't even understand all the, all these things yet. 01:17:27 - Scot Wingo Yeah. Do you, having a, you know, a nuclear background, do you track quantum computing? What's your view? I'm sure you saw the Willow Google thing and you know, they kind of had a little tuck incidents in there where it did this massive computation. And then the scientists said that we believe the only way we could have done this is if we basically stole energy from, you know, other dimensions. It's like, you know, basically he kind of like proved, you know, the, the, the multi dimension kind of theory. Like, what's your, what's your view of that? 01:17:57 - Igor Jablokov Well, I mean, look, I mean it's, you know, it's, it's, it's, it's hard for, for us to think are there other dimensions or parallel things happening and things of that sort. That's what's exciting about science and technology is that by the day we get to discover new things. So it's almost like the holidays are never ending for us. There's always something new around the corner, which is why a lot of us relish events like ces just to see what's the latest that people pop out and show us new cars and show us new tablets and new home devices and things of new cameras and what have you and just see the inevitable march of time, you know, developing new things. I will tell you though, AI is going to be a double edged sword when people like me invented, you know, somebody asked me if I felt guilty, you know, if some of these technologies became, you know, would have been something that Snowden warned folks about. And I said, you have to imagine folks like us making a hammer, but we only think of Jimmy Carter using it for Habitat for Humanity. Because most of these technologies are designed in very innocent ways. We don't think like Ted Bundy, you know, to put it in, you know, somebody's hands to start harming other people with it as well. And so inevitably there's three acts. Act one is a very innocent time for any technology. Act two is where there's malfeasance with these technologies. And then Act 3 is where people try to regulate it and then figure out how can we accentuate the positive uses of AI. What's an example where I think diseases that we once thought incurable are going to start getting knocked out one after the other. That's a Fantastic use of these style of technologies. On the negative side, you're going to see further pressures on elections. 01:19:56 - Scot Wingo Yeah. 01:19:56 - Igor Jablokov You're going to see the fact that, you know, one in three, you know, social media transactions is probably state sponsored nonsense that's generated. You know, there's an AI that will. 01:20:07 - Scot Wingo Figure there's gonna be an arms race and AI will be able to figure that out. 01:20:10 - Igor Jablokov Maybe, yeah. I mean, you have to start thinking in 21st century terms. Like for instance, you know, one of the things that I tell people to shock them out of, out of thinking, like last century is the proposition that Luigi did not kill that healthcare CEO. And they're like, what are you talking about? We found the id, we found the mask, we found the hoodie, we found the DNA and fingerprint evidence and stuff like that. I'm like, you're thinking in 20th century terms. What happened was one AI with an objective function was battling another AI with, with an objective function. The 20th century was a human being, an apex predator and a victim. In the 21st century, we're just peripherals to the AIs that are doing battle with one another. 01:21:01 - Scot Wingo Yeah. What was feeding Luigi? A, did he do it? B, what was feeding him? Information led to this conclusion. 01:21:07 - Igor Jablokov And on the flip side, you know, what was feeding that healthcare CEO? You know, Wall Street AIs telling their AI to maximize profits and things of that sort. And you literally had one objective function crashing into another one. You know, that's a new way that you have to start thinking about stuff that if you think you are craving, you know, a coffee, an orange juice, a croissant, a hamburger and things like that, are you really craving this? Or, you know, was there some form of social media or traditional media or something else that plopped that idea in your head that started, you know, you know, getting your mouth to order, as I said, these different things, or the fact that you read an article about vitamin C is important, especially now that everybody's getting cold and flus and things of that sort as well. And an orange. You started sounding good or some sort of, you know, you know, medicinal smoothie and what have you. These are all things that we can't help. They just pop in our heads. 01:22:03 - Scot Wingo Yeah, this hit me hard. I went to go on an Amazon tour to go see one of their fulfillment centers. And you go in there and there's like thousands of employees and they all have a device on their arm called a rabbit. And you know, they are all looking at the rabbit and it's basically telling them what to do. When giving them bathroom breaks and everything. And, you know, I'm sitting there, I'm like, holy cow, there's an AI brain in this guy telling thousands of people what to do, you know, and there, there's managers and stuff there, but even they're like, they've got their own rabbits. So they're not really. They're making sure the, you know, what the AI overlord tells them to do is like being propagated down. Yeah, it's like a. So very. I think maybe being a comic book, sci fi person helps you kind of think through this stuff a little bit. Greases the skids for you. Seeing that we're going to be in a world like that, it's kind of crazy. 01:22:47 - Igor Jablokov That's what's wonderful about being, you know, engineers and things of that sort. In some ways you have to think about this, this connectivity. Religion inspires art and art inspires science in some ways. You know, hey, why do we all exist? And then other folks end up creating depictions, right? Pictorial descriptions, you know, frescoes, statues and things of that sort. And then the scientists look at that and say, no, there could be these known processes that actually cause, you know, evolution or creation to actually exist as well. My parents are both artists, so same thing. For many of us, as we were kids, you know, we intersected with sci fi and most people use these things as entertainment. But for some of us, we got quiet, you know, when, when. When we intersected with. With these works and said, I think I know how to build that. It's not this fiction, you know, and. And I'll start working on it as well. And that's why, you know, we have these things that we have. I remember as part of the last startup, I got invited in Hollywood to sit down with the founder of Evernote and also an MIT professor. And it was a lot of producers, directors and writers. And they all started quizzing us like, hey, tell us what's coming, because we'll write it into our works. Then people will think that we predicted the world, but we'll also prepare the audience for you as well. So the things that you're actually witnessing and reading and watching are not as accidental as it seems. There actually could be folks that are advising each of these works, and that's why they appear so prescient. 01:24:35 - Scot Wingo Anything in our last couple minutes you want to. So for other founders out there, anything you want to say or. About the triangle region? 01:24:43 - Igor Jablokov No, I think that this is still going to be a special environment for AI. This is what I tell everybody outside of this market because, you know, is this Apple campus ramps up, it's going to have a heavy bias towards AI. And of course there's IBM, Lenovo, Cisco, sas and all the things happening here in the life sciences arena as well. So I still think it's going to have a pretty strong proposition going forward. You know, as, you know, folks amplify their investments. The fact that we can host data centers, the fact that we have, you know, pretty phenomenal utility network here as well, proximity to all the things that matter. I think good state government, you know, that supports a lot of these initiatives as well. They obviously can. There's certain neat ideas that I was on the governor's transition team that we can end up doing on, on the commerce and IT side as well to accelerate things in terms of, you know, taking a look at angel tax credit, the incentives that could be shared coming from the data centers that are getting built here to give economic advantage to locally headquartered startups, you know, you know, thinking about what is a startup and North Carolina package look like, you know, compared to what New York did. 01:26:12 - Scot Wingo Yeah, we need endowments and pension funds to invest more in the area too. Yeah, A lot of times Silicon Valley people come to town, they're actually here to grab our, you know, our big dollars and take them back out. Lust. Yeah, I would love for more of that to stick around. 01:26:25 - Igor Jablokov Yeah. 01:26:25 - Scot Wingo I don't understand. I'm not a political person, but I can at least see that happening. 01:26:29 - Igor Jablokov Yeah, yeah, there's. There's a lot of things that we can. But the bones are good, let's put it that way. It's just finishing it off and we're not far away from this stuff as well. But I do want more incentives, especially to smooth over the earliest days that are very risky. Like how do you get somebody to move from Florida up here where they're used to paying zero, zero tax. They can't even fathom that move. Or folks from Texas coming here as well. If these data centers want to plop themselves here, why as part of the economic development packages, didn't we say hey, a portion of some of the GPUs that you're going to be installing and hand it, you know, to our, our startups, you know, so that they can get some sort of advantage to get started instead of us having to, you know, you know, you know, do the March of Dimes, you know, on, on those style require a lot of. 01:27:27 - Scot Wingo Obviously you have a bunch of GPUs, but you're. Are you using more like hosted kind of stuff or do you guys have. 01:27:33 - Igor Jablokov A your own hosted for now? We used far less than than you would have expected because of how energy efficient we we were and we're going to use far more than you expect, you know, going forward because of the resources. 01:27:51 - Scot Wingo Are you heavy inference or heavy training or pretty evenly split or don't want to say both. 01:27:56 - Igor Jablokov Yeah, we do both. Yeah but the inference stuff is compute efficient enough where it can do x86 inference. 01:28:02 - Scot Wingo A lot of people are using the tpus and a variety of other like the company Grok has a device that seems like it's less GPU required Nvidia like heavy iron kind of thing. 01:28:15 - Igor Jablokov But you know, Dell, you know we're behind the scenes on Dell, we're behind the scenes on Nvidia, World Economic Forum, Air Force and things of that sort as well. We're in a lot more places than people expect us to. That's because the majority of this is white labeled where we're just behind the scenes. Which is again something that I learned from the last time. You know, for, for me it doesn't really matter, you know, whether it's our brand forward or somebody else's. You know, you just want to get that network effect where it just gets as wide as possible. 01:28:46 - Scot Wingo Yeah, Game theory tells you to build quiet and underground for as long as you can because you take less incoming, incoming attacks from competitors and stuff. 01:28:54 - Igor Jablokov Yep, yep. That's what's going to end up happening for us the next couple years and it'll be special if we can survive and represent our 10 year anniversary in a couple years. Most companies don't make it that long which is amazing. Most software companies, I think only 10% of them, or is it 15% make 10 million plus ARR after year 10. So we're going to hit that number obviously early and it's just, you know, I told everybody to and I know, you know we chit chat about about this before we started recording I told my team to prepare for 10 years of obscurity. Nobody knew what Docusign was, what Salesforce was, what Okta was, what Cloudflare was for a solid decade plus and then it ended up becoming, you know, a native part of the enterprise software stack, you know, part of that ecosystem. Same thing for prime. You know, we'll just be invisible and your friends and family members would snicker that you work here instead of Google and all these other places and things of that sort. But you know, after a decade plus then people will know what we're about. 01:30:04 - Scot Wingo Cool. And personally hoping, hoping for a big IPO. I think we need more big, splashy IPOs in the triangle. 01:30:10 - Igor Jablokov So, yeah, we're going to definitely have the optionality for that. Part of that is hiring senior executive executives that have been there, done that before. Yeah. So not us back a proper IPO and, And things of that sort. You always kind of want to have a dual path where you, you, you lead the possibility of a. Of a healthy M and A or an IPO and just. And essentially build for that optionality. Just building for growth. 01:30:36 - Scot Wingo Yeah. Yeah. I'm cheering for ipo, but obviously you got to pick the best path for your. 01:30:40 - Igor Jablokov Yep. Whatever. Whatever we get allowed to. Yeah. 01:30:44 - Scot Wingo We've been very generous with your time. We appreciate it. 01:30:45 - Igor Jablokov Yeah. 01:30:46 - Scot Wingo And thanks for sharing your insights with our audience and thanks for doing this. 01:30:49 - Igor Jablokov Yeah, thanks for having me. For more Tweener content, check out the Triangle Tweener time substack@tweener.substack.com. for more tweener content, check out tweenertime. Thanks for listening and we'll see you again soon on Triangle Tweenertalks.