This Week in Startups - TWiST News: Venture-Backed Defense Startups, Anti-Drone Guns, and Querio's AI Data Platform | E2048

Starting point is 00:00:00 All right, everybody, welcome back to this week and startups. I'm Jason Kellelicanis, an angel investor and podcast hosts here in Austin, Texas. I used to say Silicon Valley last year. Now I say lots in Texas. And with me, my co-host, Alex Wilhelm, how are you, sir? I am fantastic. We have an amazing rundown today. Yes, we do.

Starting point is 00:00:20 And we also have Nvidia earnings after the bell. That's going to come out, of course, after we record this, Jason. But I am just bursting with excitement. I mean, this is going to be, I think, earnings. that really sets the tone for the next couple of months. He's going to tell people where we are on the bullishness hype cycle for AI. So I'm kind of counting down the minutes right now until Jensen tells us if there's going to be Christmas in technology or just coal in our stockings.

Starting point is 00:00:45 This weekend startups is brought to you by Vanta. Compliance and security shouldn't be a deal breaker for startups to win new business. Vanta makes it easy for companies to get a stock to report fast. Twist listeners can get $1,000 off for a limited time at vanta.com slash twist. Open phone. Create business phone numbers for you and your team that work through an app on your smartphone or desktop. Twist listeners can get an extra 20% off any plan for your first six months at openphone.com slash twist. And Sprig, the product experience platform that generates AI-powered opportunities to continuously improve your product at scale.

Starting point is 00:01:26 Visit sprig.com slash twist to book a demo and get a $75 gift card. Let's just get into it. They're expected to report $33 billion in revenue. Yeah, if you put that number times four, you get to a $130 billion a year. And the data center revenue is $29 billion of that. If you didn't know, Nvidia used to be a company that made graphic cards

Starting point is 00:01:56 for gaming. That was basically their business. That's what they were known for. If you were a video game nerd, you would sit there and you'd buy an invidia card if you wanted to get your frame rate up when you were playing a desktop game. Somewhere along the line, people took those GPUs, graphical processing units and started using them for cryptocurrency. So then they had two line items and cryptocurrency, I think, for a time, you know, was driving the stock. Those two little businesses. That was their there, TAM, total addressable market. And we're going to talk about TAM a lot, because I just saw a great clip of Bill Gurley talking about TAM. I think maybe we'll put that for tomorrow. But that's an industry term for, hey, what's the market size you're going after? And you can do all kinds of fun things once you know the TAM. The TAM of gaming and the TAM of Crypto,

Starting point is 00:02:48 well, you can kind of figure out what the total addressable market is, how many customers there are, how much of they're willing to spend. and then you can do interesting things, like who has a certain percentage of the market, and then how much is the market growing year over year, and that can give you some ideas of where the company's going. Why is this so flawed, Alex, when it comes to looking at Nvidia at that time? I know we're going to get to the Uber example tomorrow,

Starting point is 00:03:17 but I'll just kind of tease it for everybody. But one way to think about Uber back in the day, but before it was big, when it was small, was to look at the taxi market and say, okay, the taxi market is X dollars. Uber will get 30% of it. So its revenue ceiling is, Y dollars. The thing is, a really good product makes the market much larger. Uber made the taxi market larger. iPhone made the smartphone market larger, et cetera. So if you have a technology that breaks out, for example, Nvidia GPUs in the data center context, you can suddenly go from a company doing a couple of billion a quarter to dozens of billions a quarter.

Starting point is 00:03:51 Right. And I call this market manifesto. station. And I got that term from induced traffic. I remember when I lived in Los Angeles, they would always be adding a lane to a highway. They'd be like, hey, you know what, the 405, it's getting a little congested here at the 405 and the 10. Let's add some bigger off ramps. And they did these spectacular off ramps. And I remember reading about it, Alex, these were going to change everything. These off ramps, instead of one lane getting off in a tight term. turn, they were going to have two lanes and they were going to be a giant wide turn onto Wilshire Boulevard or Olympic or Santa Monica Boulevard. So when you got off the 405, instead of having

Starting point is 00:04:36 this little quarter mile circle of one lane with 10 cars in it, let's say 20 cars in it, you would have a mile with 40 cars in two lanes. You'd have 80 cars getting off instead of 20. You went from 20 cars getting up to 80. And that would make the traffic flow because you could get more cars off quicker into that queue to get on to Santa Monica Boulevard. You know what happened? Yeah. But people started to realize, hey, traffic on the four or five has been easy, breezy. Maybe I'll buy a house a little deeper into the valley, get a little more bang for my buck.

Starting point is 00:05:11 Because, hey, the 40 minute commute, I can go an extra exit or two out and it's kind of reasonable. So I'll get the cheaper house. Or, hey, you know what? I love this restaurant in Santa Monica. I live in the valley. or I'm in Venice and I want to go to a restaurant in Clover City. Yeah, I'll zip across and go to that restaurant. It would induce more traffic and it never ended.

Starting point is 00:05:31 So that's what Uber did. In this example, though, Nvidia became the choice for doing large language models and inference, I guess, on the margins as well. And these H-100 servers just started doing bigger and bigger machine learning tasks. Large language models come out and suddenly everybody, the tech industry sitting on mountains of cash realizes, hey, we can't buy anything. So here's the other trend, right? The wrath of Khan, which is coming to an end. The wrath of Lena Khan said no M&A,

Starting point is 00:06:05 that means all those cash builds up. Nobody's buying whole foods. Nobody's buying. You know, what's that? So what are you going to do with the money? You might as well build some servers. You might as well stand up a server farm. It's a good use for it. Maybe a stock buyback. And that's when $29 of the $33 billion that Nvidia is expected to announce tonight is going to come from data centers. They just let that sink in. 90% of their revenue now comes from a product or a business line, which was not really on the top of people's minds, but five to 10 years ago, correct?

Starting point is 00:06:39 Oh, absolutely. And Corey, can we get that table back up on the screen for a second? Because I want to make a point to underscore what Jason is saying here, which is that the scale of Nvidia's revenue, I think, kind of occludes or hides how quickly it came to be, because you hear that number, you think, okay, well, how much revenue does Apple do or does Microsoft do, whatever? But Jason, if you look at this chart,

Starting point is 00:06:58 just observe the jump from the first quarter of their fiscal 24 to the second quarter of the fiscal 24. People are not watching the video. Nvidia's Data Center revenue went from 4.3 billion to 10.3 billion in a single quarter. And then it went to 14.5, 18 and a half, 22 and a half, 27. and then today expect to be 33.

Starting point is 00:07:19 That is still accelerating, Jason. It's still getting much bigger very quickly. And that's why I think this earnings report matters so much. There's a lot of enthusiasm right now in the market in a post-Trump context. The Lena Con tenure is coming to an end. There's still a lot of investment hype and a lot of excitement about what's going to come. The vibes are good in other words. Yeah.

Starting point is 00:07:38 And I think that Invidia will either maintain that or really hurt it. So give me the case for hurting the vibes. what could cause you know the stock to tank what could cause people to lose faith what would be the narrative or the plot lines

Starting point is 00:08:00 or the vibes that could make people go you know what maybe this is overhyped and coming to an end yeah so I think two things there one if the company just meets expectations like I believe we saw in the preceding quarter we could see a stock drop five six points again. People are expecting a beat. This is a company that is valued on growth.

Starting point is 00:08:20 It's expensive by a couple of traditional metrics. People really believe in it, so it's being valued as such. The reason why I think it could harm the vibes if it misses or just barely meets expectations is that that implies quite a lot about the rest of the industry. Invidias, quarter to quarter revenue numbers, Jason, are a proxy for the investment that big tech companies, the hypers, are putting into their data centers, which are a proxy for market demand for AI models, AI inference, and essentially just the health of the AI economy. So, Nvidia's kind of down the pipe a little bit, but it should tell us what's going on upstream, and it's one of the best indicators we have. So that's why I think it matters so much for the software picture,

Starting point is 00:08:59 if that makes sense. Founders, do you want to sell to bigger customers? I know you do. You got to get that ACV trending up. And you want to push your churn down, right? Sounds good. But to sell to those big buyers, you need to clear all of these compliance checks. You know that. That means you got to have things like SOC2 sorted out. What's SOC2? It's a standard and assures that companies keep their customer data safe. And if you aren't SOC2 compliant, you can kiss those big deals goodbye.

Starting point is 00:09:28 You're not going to land the lighthouse customers. You're not going to be able to operate at the highest end of the market, but Vanta makes it really easy for you. To get and renew your SOC2 compliance, on average, Vanta customers are compliant in just two to four weeks. It can take months without Vanta. and they automate compliance for GDPR, HIPAA, and more. So you could sell to bigger customers in whatever markup your startup is going after. Vanta is going to save you hundreds of hours of work and up to 85% on compliance costs.

Starting point is 00:09:56 Stop slowing your sales, team down, and use Vanta. Get $1,000 off at vanta.com slash twist. That's vanta.com slash twist for $1,000 off your sock, too. This quarter, are they on a fiscal quarter or are they on a calendar quarter? I'm curious. This is Q3 fiscal 2025 or 2024? Am I reading this? Yeah, so this is, we're in their fiscal

Starting point is 00:10:19 we're in their next fiscal year, a bit like Microsoft. They're offset by a couple of quarters. Got it. Okay. So this is going to be really interesting. We've never seen a company grow large amounts of revenue in this way. The last quarter was $30 billion. I'll give me if I'm saying that. And the year earlier for

Starting point is 00:10:41 Q2 was 13 billion. So they more than two times their revenue that quarter to that quarter. Q3 last year, 18 billion, this year 32. So it's going to just under double, right? So this crazy growth is slowing, but it's still significant. And we talked about before on the program, there have been a bunch of people looking at round-tripping as a technical era, as a technical term for self-dealing or insider re-transactions. They invest in a company, that company buys servers with Nvidia's money, but those numbers were kind of small. Then there's

Starting point is 00:11:24 accounting issues. When do you recognize revenue? So a lot of the revenue here, my understanding, is baked in because somebody will put it in order. You have a company like X.A.I. Elon's X company, they bought a lot of H-100s. They stood them up incredibly quickly. Okay, those all got delivered in a quarter, but they probably have future ones coming in. And then you have people like Amazon, Apple, Google, and meta, probably programmatically buying X number per month. My understanding is Jensen's approaches, we just take your order and we fulfill orders and

Starting point is 00:12:00 me when orders are received. But I don't know how the money works. Do they get the money ahead of time and they get to sit on it? It would seem like they have the unique ability in the world to dictate terms. And I wonder if, you know, how much cash they're sitting on now and what they require in terms of people giving them money in advance to lock in those. One could argue that they could give people two options. One, you put a 10% down payment on. You're in the queue.

Starting point is 00:12:30 You want to be in the VIP queue. You put down a full deposit, 100%, and you go to that. that. So I would run the company with two cues that, you know, whatever, 25% down queue, and then you get stuff after that, and then the priority key, which is you pay 100% up front. I don't know exactly how they do it. So I'm looking through their last quarter's earnings, because again, people were recording this right before the Q3 fiscal 25 numbers. But if we'd look at their deferred revenue, which Jason, I believe is where they would put prepaid contracts, essentially, it's not that much money.

Starting point is 00:13:07 It's actually, at the end of the last quarter, just $1.7 billion worth of deferred revenue. So to me, that doesn't imply that there's a lot of chicanery going on here. And I'll just point out that the investing community, court, if we could have that table pulled up real quick, the investing community is not concerned about this. As you can see from here,

Starting point is 00:13:28 Nvidia is once again, the most valuable company in the world, beating Apple, beating Microsoft, beating alphabet, beating Amazon, beating the Ramco, beating meta. So while you and I, because we're nerds for financial documents, are a little concerned about possibilities of round-shipping or what to do with the VC fund, as you mentioned last time, the street doesn't care. Great. Awesome. Okay.

Starting point is 00:13:52 So we are going to get those numbers and we'll have a live show tomorrow Thursday. Yes. Because instead of doing a Friday show, I'm going to be taping all in on Friday. So you get your all in on Saturday. but we moved up this week and started up to Thursday so I don't have to tape two episodes in one day which is exhausting for me one a day is great it's like a workout

Starting point is 00:14:11 two a day too much whenever I do two records in one day I walk into the house afterwards and my wife's like what's wrong and I'm like I had to talk for three hours and she's like boohoo you poor thing but it is it's very tiring well it's mentally

Starting point is 00:14:26 stimulating you and I get to have a conversation here but all in sometimes tips in from just a conversation to just outright sparring and a cage match, especially if things like Ukraine come up or Trump or whatever political discussion. So for me, this is like a delightful conversation. That can become a little bit, yeah, spicy, let's say. And then try, on that one, I'm the moderator. Here, you get to play a little bit more of the moderation role. And I get to kind of shoot a little bit more here to your point guard and we do the pick and roll over there. I'm doing a little more point guarding, which is also exhausting.

Starting point is 00:15:04 Yes, it's much harder. People have asked me, do you prefer to be a panelist or a moderator? And the answer is so obvious. Being a panelist is great. You show up with your shoes untied. You sit down. Someone puts a mic in your ear. Right.

Starting point is 00:15:17 You shoot. Right. Easy. They pass you the ball, you shoot. Yeah. Man, coming down the court, having two guys double team you, and you're trying to zip around and get somebody open and you have to keep your entire peripheral vision open. And hey, where are the opportunities here for other people to score?

Starting point is 00:15:33 Takes a lot of work. All right. So that's Nvidia. Yeah. Let's move on. Anything else we should be looking for there or anything notable or just we'll talk about it tomorrow when you can see. We'll talk about it tomorrow.

Starting point is 00:15:43 I think that's enough on that because we'll get the numbers. We'll do a better deep dive. But that's the state of play for 4 p.m. today for on the last term. Okay. Jason. By the way, the number one thing I want to know just as a project here is when does Nvidia have a competitor? This is my big question.

Starting point is 00:16:00 When do we think there'll be a disruptive competitor and you'll see a leading, this is what I'm looking for. Okay. A leading company, a leading customer, one of their lighthouse customers, let me say it that way. When will one of Nvidia's lighthouse customers, that's a technical term in our industry, for a customer so pronounced that other people are guided to your product because of it? Like a lighthouse guides people to shore safely, having XAI or having open AI or having Amazon Web Services as your customer that would guide other people to the safety of an H100 in the NVIDIA's term. So when does a lighthouse customer flip to another product? Okay.

Starting point is 00:16:43 That's what I'm looking for. That will be the disruptive moment we should all be looking for with NVIDIA. And it will come first in the in the form of an announcement. then there'll be a startup that tries it and gets a lot of attention for doing it and then there'll be a competing cloud computing offering. So look for that little cohort.

Starting point is 00:17:03 That's when you know the invidia story is going to have complications. Yeah. Well, one, I mean, just to pick a $2,500 company etched, they're making chips that are literally purpose built for the transformer architecture for LLMs, which is a wager,

Starting point is 00:17:19 by the way, on that maintaining its primacy in how AI models are built. But I love it because if they're correct, they're going to be a huge company and could, to your point, Jason, snag some real market share. I don't think we'll see hyperscalers stop buying Nvidia chips altogether, but I so I would amend your what to look for by saying, when does the, when does a lighthouse customer for Nvidia start to buy large amounts of a competing chip? And of course, Amazon has its training two chips coming out. So there is some other names on the agenda. Jason, before we get to a guest, though, I want to talk about AI training

Starting point is 00:17:53 and data because there's a little bit of news here from the Twist 500, and it's that I don't think we actually got to this on the show, but Tollbit, one of two companies that we added that are dealing with building a marketplace between content providers and AI models that want to use that data, raised a Series A, a $24 million series A quite large,

Starting point is 00:18:12 and they said as part of that they have customers, they have data, and they have AI companies on the platform. So progress there, pretty good. And then also pro rata.a.a.ai, a new company to me, backed by Mayfield, just secured a $130 million evaluation

Starting point is 00:18:26 because it's doing the same thing over in the UK. Dot AI. Yes. Prerata being your share of something and your ability to buy those shares in the future, that's the term we use here in the Valley. So fair compensation and credit

Starting point is 00:18:43 for content owners in the age of AI. So this is not pro rata as in you get, if you own 10% of a startup, you get to buy 10% of the next round. This is compensation for content orders. This is amazing. We need to have attribution.

Starting point is 00:18:57 We need to have citations. Exactly, as I had said on this very program, I think two years ago, at some point, citations will be required and permission will be required. Citations are in the latest version of chat GPT everywhere. Perplexity has always had some of them. And in the new search product that ChatGPT, or I'm sorry, Open Eye is doing, their search feature. have some citation. Some of them are buried, some aren't. So I think this is what we're seeing

Starting point is 00:19:27 is the healthy evolution of the ecosystem. And what's going to have to happen is the untraining of models. So if you put Reddit in your model, when you build your next version, you're going to need to make sure that all that content from Reddit is taken out somehow. Technically, that might mean, you know, yeah, I mean, how do you do that? It's a, it's a technical question. I'm not sure how much, how much of a new version of chat GPT relies on what was indexed previously. And so, yeah, I don't know the answer to that. I know in some models, they just start over, they have all the data sitting there. So then it would just be like, it would be the equivalent of having a library and going in and saying, okay, take all the Stephen King books down, take all.

Starting point is 00:20:15 all the token ones down. We don't have the rights to those. Take them out of the library. Now open the library to customers. But if all the knowledge was already sucked in and it's got it in there, I don't know how you rip out that scaffolding. That's a really good question for when we have these companies on the pod. One thing I'd like to institute here, if somebody is good enough to be on the Twist 500,

Starting point is 00:20:36 let's have the founder on the pod for a quick guest hit. Twist 500 is our attempt to identify the top 500 private market. companies. Now, an easy way to do that would be to just look at market caps and put in all the unicorns. We're not doing that. There, there, you can sort the Twist 500, which is built on Coda. Yes, it is. I.O. slash twist, I think, and you get some sort of free deal. We didn't pick them because they're a sponsor on this podcast. Um, we did because it's a really great tool for doing stuff like this, but it's a database. You can go in there. How many companies are we at right now? We are about 110. And, uh, due to a snafu, I'm re-adding it about another 10.

Starting point is 00:21:15 so we'll be about 120. All right, going slower than I would like, but maybe this will be our Q4 end of the year project is to ramp up here. Somebody had a baby and I think has been having a little less sleep. We've also been doing a lot of shows. That's true. We have been hitting four or five shows in a week,

Starting point is 00:21:34 which I never thought was going to happen. But next year it's going to be three shows a week. The sales team hates me because I'm turning down money for that fourth, fifth show a week. But I just, you know, I want to do three really high quality shows a week. Let's get that dialed in. And then maybe we can both get some sleep.

Starting point is 00:21:54 All right, let's, you know, we have a, speaking of guests. Yes. And you can check out twist 500, twist 500.com. Go ahead check it out. There's a submission form there. Don't email the sales team and ask them to put you on the twist 500. People selling the ads. If you go through them, it's probably going to be worse for you.

Starting point is 00:22:10 That will be like a minus one because we have a Chinese wall, a far. between the editorial and the ads. You don't get to be on the twist 500 because you advertise. Okay. But there is a submission form. And I don't think anybody reads it, but there's a submission. It goes to one of my email addresses, I believe, so I will see it. Founders know that every missed call is a missed opportunity.

Starting point is 00:22:32 Customers don't want to wait. They will call someone else if you don't pick up. But if you use open phone, you're never going to miss another customer call. And guess what? It's super affordable and easy to use. used to spend tens of thousands, hundreds of thousands of dollars, putting in a corporate phone system, and now for just $15 a month, Open Phone will give you a business phone line and complete control of your destiny. Want to know who's answering customer calls and how they were handled?

Starting point is 00:23:00 Open Phone can do incredible things like sync with HubSpot and give you AI powered call summaries. Automated responses to ensure you don't miss a single ring. Well, that's all built into open phone. And if you've got an existing phone number, they will let you port it all over. at no extra charge. Get 20% off your first six months. What an amazing offer at openphone.com slash twist. That's O-P-E-H-O-N-E dot com slash twist for 20% off for six months. Let's talk about our first guest today.

Starting point is 00:23:30 Sure. There's a company, Jason, called Allen Control Systems. Oh, yes. They caught my eye on Reddit because I like things that go boom. And then it turned out that they are a venture-back startup and raise money, I believe, from KraftVentures. So I thought, hell, let's have them on the show. Destiny David Sacks back this company. I remember we had a conversation on All In last year, or maybe two years ago, we were talking

Starting point is 00:23:50 about would you back a company? And I remember Sachs kind of waffled a little bit at the answer of the question. And then I found out he actually had, but he hadn't announced it yet. And Steve is here. Welcome to the program, Steve. Hey, thanks for having me on, guys. Appreciate it. All right.

Starting point is 00:24:07 I think you might remember that moment, maybe, because... Yeah, I do remember the moment. Also, it's good to see you again, Jason. and I briefly saw you at the All In Summit. I was thinking of mostly Jimmy, though, your guy, Jimmy. Oh, Jimmy Donahue, my, my, my, my, we're going to golf together. Are you really? Jimmy D is the greatest.

Starting point is 00:24:25 Jimmy D. and I grew up on the same block in Brooklyn. We had the same crew. He went on with my brother to join the NYPD. The three of us were going to go on at the same time. Right as I was about to join NYPD, I got accepted to Fordham at the last minute. and I went at night. And there is where like the changing cars, the butterfly effect, whatever, happened.

Starting point is 00:24:51 Yeah, Jimmy's the man. Jimmy is the guy and he does all my security and is awesome. So tell us a little bit about your company. What are you trying to do here? Yeah, I mean, fundamentally, we're trying to change the battlefield economics. So Allen Control Systems, we make autonomous gun systems. And so these are guns that can point and shoot on their own. So the operator, the soldier only has to give a command of what target they want.

Starting point is 00:25:18 And then Bullfrog, as you see here, will actually do the rest. It'll slew to the target and then fire the shot. Wow. And it is an actual assault rifle, it looks like. That's been mounted. So you don't make the gun. You use an existing gun in the world, correct? Yeah.

Starting point is 00:25:38 So we're using a standard M240. assault rifle. It's a standard, it's a standard Army gun that's in, you know, most units. It's a very widely distributed gun. Yeah. Got it. So you don't have to worry about the maintenance of those guns or building them. They exist in the world already. Is that a 50 caliber on this one I'm seeing here? That's still the M240. We do have a 50 cal variant coming. So what we make is the, we make all the steel, like we cut all the steel. We make all the circuit, the motherboard, the motor control board, all the circuit cards are custom and then the software. Everything about the gun, basically. Well, because you know, when I did my tours, Alex, stolen valor, I used to man the 50.

Starting point is 00:26:15 In my dreams, I man the 50. My, my, my, my, my, my friend. You're like Cuba Gooding Jr. in that one movie, Pearl Harbor, man in the 50, Cal. Yeah, I used to man that 50. I used to man that 50. I think also Blackhawk down there, were yelling man the 50. And John McEughan, my good friend who was a green beret, he used to man the 50 caliber. All right.

Starting point is 00:26:33 So you have this existing assault rifle. You build a container for it. and then it has a lot of AI. And this is not used to go out and kill humans in the field. This is used for defensive purposes today, correct? Yeah, so we are starting with the hardest problem, which is hitting a fast-moving small drone. Obviously, it is a gun system.

Starting point is 00:26:56 So, you know, the Army will use it for multiple mission sets. But drone warfare is the primary concern right now. I'm really curious about just the sheer volume of these because I've actually shot machine guns before, once at West Point, for example, and they are just so indescribably loud. And so I'm kind of curious, does that at all impact how you design systems

Starting point is 00:27:17 go after drones? Because I presume you want to be stealthy until you pull the trigger, but oh my God, once you do, Stephen, they're just insane. Yeah, that's a good point. So our system is designed to be passive,

Starting point is 00:27:27 meaning it's using cameras to find the drones. So instead of a radar detection system, which is very loud, the enemy can see someone using active radar, and then they can hit an artillery shell to your truck. Our system with just the cameras, you know, before you fire the shot, you can hide. And so the passive nature of it is really good for force protection.

Starting point is 00:27:46 Okay. And then does that mean that it's optical based? And if so, does that mean that you have issues in rain, snow, sleep? Is this a fair weather system, I suppose? Yeah, I know. It should work in all weather. And at nighttime, we use EO, electrical, optical, and IR sensors, so infrared. And so we're able to see it night as well.

Starting point is 00:28:02 You know, obviously, if it's really rainy out, that can be a problem. But that's also a problem for many other weapon systems as well. Let me ask you a candid question here. How long did it take to develop this system just in quarters or months or years? I mean, this was a solid 18 months to get it to where the government is buying it now. I would say it's another 18 months, you know, until it's like the best gun on the market that you could buy. Like it's going to be, it's already the most accurate gun in the U.S. right now. But it will be incredible, you know, in another year or two.

Starting point is 00:28:34 but yeah, like right now 18 months. Got it. So the reason I asked that question is there are bad guys in the world. And if you're a small team at a startup can build this in 18 months, is there, are there versions of this that have been built with by nefarious players, terrorist organizations, bad actors in the world that you have become aware of since starting the company? Not that I'm aware of. This is the, I do survey the market pretty often as the startup CEO.

Starting point is 00:29:03 This is the first gun that is putting a one or two couple bullets on a drone at any like two or three football fields away. So there's nothing like this on the market. There will be, I mean, obviously anyone can build anything. This does take a lot of money. Obviously, I was fortunate enough to sell my last company

Starting point is 00:29:19 to DoorDash, which helped fund a lot of the initial. It was called B-Bot. It was an online ordering point-of-sale company. Oh, cool. Actually backed by Kraft Ventures, so it made them some money when we sold that for a good amount of money. So you went from Point of Sale to pinpoint

Starting point is 00:29:35 a sniping. Well, before the restaurant technology, I started my career in the U.S. Navy. I was a U.S. Naval officer. And actually Luke Allen,

Starting point is 00:29:46 my co-founder, the company is called Allen Control Systems. I named it after my friend. I met him in the Navy. So we worked in the nuclear, it's called Naval Reactors, NAV-C-O-8.

Starting point is 00:29:56 So we were the headquarters for the nuclear Navy. Oh, wow. That's awesome. So the, I know, submarines are powered by small nuclear reactors. Yes.

Starting point is 00:30:07 Yes. So we would train all. Are the aircraft characters? Yeah. Yeah. So we saw, we oversea, saw from cradle to grave,

Starting point is 00:30:14 the nuclear reactors on the carriers and the submarines, and we trained the fleet on how to operate. May I ask a dumb question to, Steve? We have big debates about small modular reactors and safety of them. Our military, there are dozens of small. nuclear reactors, I believe, on submarines and aircraft carriers, correct?

Starting point is 00:30:40 Yes. Yeah. So there's dozens. These are very powerful, you know, reactors on the submarine. Small in the sense of volume. They are not the size of the giant, you know, spherical cones that we see on the landscape. These are the size of a conference room or 10 conference rooms. Probably like the size of, you know, like eight conference rooms.

Starting point is 00:31:02 Okay. So it's the size of a couple of tractor trailers inside one of these larger things. And they are safe. Has there ever been an instance where one of these dozens of them have had a meltdown and are not safe? Naval reactors, we would say we've had in our, I guess, 75 years of operation, no reactor accidents. So when you hear people, you know, hand-winging, kovetching, whatever, about nuclear reactors, do you just pound your head into your desk? I, well, I'm a strong proponent of large scale commercial reactors.

Starting point is 00:31:36 I think those are the best. I think we need more of those. The small modular reactors, you know, while they look promising, there's a lot of startups doing it. It seems safe. It's just, it's not where I would want to be. Like, if I was investing my money, I want to be in the big reactors. Got it.

Starting point is 00:31:48 Economies of scale? Economy of scale is just super efficient. They can pump out a lot of power. Like, people are just afraid of them from a, it's a public perception thing. Public is a little afraid of them, but you can put these in faraway places and, you know, run the power. So I'm a big reactor guy. The small modry reactors are still, I still think it's to be seen how they play out.

Starting point is 00:32:08 That's where I'm, that's where I stand. All right, Stephen, I want to get back to drones, though, because we've been talking about drones. I can talk about any topic, by the way. We can go, I do my podcast. You're a good guest. I got it. All right. So, first of all, opinions on the new O-Path album.

Starting point is 00:32:22 And do you think they'll return to the real death metal roots of the band? Or do you think they're going to stay progier and more melodic? All right, you got me. You got me. You can't comment on death metal. No, you got me. Damn it, Steve. Everyone should be able to come.

Starting point is 00:32:34 I understand. So that's me. Good job. All right. You want to build products that users love. And we all know, product market fit is how you know you've built something special. And if you're hunting for product market fit, you need to understand your users better, right? How else could you boost engagement and drive more conversions for your product?

Starting point is 00:32:56 Well, product analytics alone is not going to get you there. So let me tell you. you about Sprigg. Sprigg is a product experience platform that generates AI powered opportunities to continuously improve your product at scale. Let me explain to you how this works. Sprig captures your product experience in real time, and then they use AI to instantly analyze your product and give you real-time insights. This is going to let you know how to make your product better. And if you make your product better, it's going to drive revenue, it's going to increase retention, it's going to reduce churn, all of those good things are going to happen. See why top product teams,

Starting point is 00:33:31 that Notion and Figma, two of the classic breakout products of the last decade, are already using Spriggs AI to unlock new opportunities at scale. Visit sprig.com slash twist to book a demo and get a $75 gift card. How nice. Sprig, S-P-R-I-G dot com slash twist. Steven, um, drone swarms. So we've talked to Skydeo.

Starting point is 00:33:56 We've talked to other people that are working on both in the air and underwater drones. and they're talking all about swarms. And when I look at your system, super cool, love the technology, seems to be able to hit single targets. Is it applicable to drone swarm defense, or is it more we're going to get, you know,

Starting point is 00:34:12 spare reconnaissance's drone that's flying over and knock out? Yeah, this is meant for a drone swarm. So it'll do coordinated fires, so multiple bullfrogs. The bullfrog itself can turn 360 degrees and point straight up in the air. So imagine if it's on it, if the bullfrog is on a Bradley tank,

Starting point is 00:34:27 it can do a full protective bubble around the tank. And it shoots very quickly. So, you know, we think we can shoot with one bullfrog up to like 20 drones in under like 15 seconds. Like it's that fast. Yeah, no. And so, but it's going to take multiple bullfrogs and swarms could overwhelm it. That is very true. I think that we will see like 10,000 drones coming at something. You're going to need a lot of guns and maybe some other layered defense for electronic warfare or something else like that. Well, here's the good news. these things, I'm going to guess. I'm going to guess the price here.

Starting point is 00:35:01 $500,000 per? $350,000. Okay. I'm becoming an expert on this. Firm fixed price, by the way. Firm fixed price, not cost plus. Not cost plus, exactly. So the reason I'm becoming an expert on this is because I have a podcast,

Starting point is 00:35:14 you understand. And we had another company on recently that's making out of Providence, Rhode Island, unmanned, basically torpedoes that are only 70, thousand compared to like what they would normally cost. So you're part of this new generation of instead of doing cost plus, hey, you make it, it's got a good margin in it. Maybe there's a maintenance contract or a software license to, you know, keep some reoccurring revenue come in or whatever. But the army can buy ten of these. And if you lose five, who cares? Yeah. They're disposable at 350. Exactly. It's a very good price point for them for this kind of exotic

Starting point is 00:35:51 technology. And yeah, we hope to, like I said, we're trying to bring that cost curve down. It's so important. Yeah. I want to go back to the Bradley point because Bradley's are APCs, right? Versus tanks. The Bradley is the Army tank, the big army tank. I thought that was the Abrams. Oh, yeah, the Abrams.

Starting point is 00:36:08 Well, the Abrams and the Bradley, I think, are both. Oh, anyways, I did not mean to drag us down to MBTs versus APCs. But the point is they're mobile. So if I had an infantry division, say, and I had several different wheeled or tracked vehicles, I could essentially bring full-on drone defense with me for an entire cohort of troops, presumably pretty much any climate. So does this negate the drive we've seen

Starting point is 00:36:35 towards more drone-based warfare versus human-driven warfare? I think, yeah, well, I think when our product really hits the battlefield, which will be soon, it's going to dramatically change how FPV makers think about their companies. We're about to make them look pretty dumb. Wait, what's an FB?

Starting point is 00:36:54 FPV drone, like a small first person drone. Like, we're going to... Oh, first person drones. Actually, I have one here. This right here. Oh, yeah, yeah. Those are like the standard DGI. Yeah, so like these are like, you know,

Starting point is 00:37:05 companies are trying to strap bombs to these and then fly them at artillery or Bradley's. Yeah, Bradley is like the smaller. Like the Abrams is the bigger tank, you're right, but the Bradley's like tank-like. So you send one of these with a grenade on it or something a little bit more and you drop it or you kamikaze it in, right? So exactly. And so we're going to be neutralizing that threat. And I think that the drone makers are going to have to respond some way because it's really hard to, you know, mitigate a bullet hitting this.

Starting point is 00:37:31 I was just about to say you can't put armor on that thing because one, you've put enough armor on to negate a 50-cali bullet. It's going to be too heavy to fly. Yeah. And two, I mean, it wouldn't be maneuverable. It would be just a brick in the sky. So it would be easier to shoot at that point. So is there a way for them to like see incoming bullets and then get out of the way? I mean, we're going to give the drone maker.

Starting point is 00:37:51 So there's startups like Nero's doing drone swarms And then there's Anderil just released the Bolt. Everyone was using the Bolt M. It's a, the Bolt is their, their cheap FP by Andriel. It's actually, it's incredible,

Starting point is 00:38:02 it's an incredible product. I was really like it. We're actually doing a test ACS versus Anderil next year. There are Bullfrog versus their drone swarms, their bolts. Nice. And so,

Starting point is 00:38:14 can we pay-per-view this? This would be a great this week in startups pay-per-per-view. Yeah. Here's the Bolt. Yeah. The bolt is there, Is it super sonic or no? No, it's not super sonic, but it's very fast.

Starting point is 00:38:25 And so this thing is, they're going to try to see if they can get through bullfrog and blow up like a cyber truck or something. And we'll, uh, we'll do that competition. Again, I feel really strongly about how we're going to perform because if you attach any sort of payload to bolt, it can't move as fast as it shows in the videos. Like, if there's a four kilogram or three kilogram bomb on it, like it's a sitting duck for a bullfrog, I think. Well, I think what's going to happen,

Starting point is 00:38:50 in these cases, I'm going to take a guess here, is it's going to be about, kind of like the Iron Dome in Israel is, it's just what percentage can you take down with the bullfrog? And then how many can they send out once and can some slide through? So this is just going to be, correct me if I'm wrong, how many get through the net, how many? Yeah. Yeah. Well, there's a couple of things we're doing there. Like, so our company's strategy is to make a family of autonomous guns. So Bullfrog is a small caliber, M240 based. We're also building an M-130-mini-gun, so it's a Gatlin gun. So the Gatlin gun can do like 3,000 rounds a minute. So imagine coordinated Gatlin, autonomous Gatlin guns. I mean, we're going to put a lot of lead in the air.

Starting point is 00:39:32 It will cover the sky. It is what it is. Yeah. Cheap. I mean, a bullet is cheaper than a drone, and a drone's cheaper than most other things in the military world, but we're really getting back down to kind of like dollars per kill here. Yeah. And what happened, what we really, the insight, really from Luke, my CTO, the brains behind this is that it's the control systems. That's why the company's called Allen Control Systems. We just found a way to put a bullet on a small drone at three football fields away. And it was just a very hard problem that, you know, we're happy. We're very excited right now.

Starting point is 00:40:00 You're going to put these on sniper rifles, I take it at some point to get people further away? So at the longer range is what we're doing is a 30 millimeter chain gun. So it's like a, it's the Bushmaster by Northrop Grumman. My dad actually worked at Northrop Grumman for 33 years. So I grew up around this into the prime contractor world. How many people in your company right now? So we're 35 people adding about four engineers a month right now. Yeah, just grown.

Starting point is 00:40:25 So what does this say about our military and our ability to compete in the world when we take capitalism, entrepreneurship, and unleash it on the military industrial complex? What does this mean for our ability to be? the most important military in the world. Yeah, I think this is actually a really, I think about this a lot, this question, because if you didn't have the background me and Luke had and then you hadn't sold a company for a lot of money,

Starting point is 00:41:00 we wouldn't have been able to get the capital to build this product. And I think this is a huge problem in the U.S. right now. So there's a lot of smarter people than even me and Luke that are in universities that have great ideas like Bullfrog, but they can't even really get started to like help the military with these new innovations because the capital constraints to get in, the amount of capex you need is so high. So it kind of,

Starting point is 00:41:20 I think we're missing out on a lot of new type of weapon systems that we should be building. And the government's not capital allocating into like new innovation like that that much. So like without, I don't really know what the U.S. would have done without like the situation

Starting point is 00:41:32 Luke and I just found ourselves in to able to do this. Well, we do have the venture industrial complex. And I think the venture industrial complex combined with entrepreneurs, taking on the military industrial complex, or at least creating a little bit of competition on the margins is going to be transformative.

Starting point is 00:41:54 And if you look at any top-down government, dictatorships I'm referring to, they can move quickly because they can enslave a million people, put them into factories like the Chinese are doing with Uyghurs, and say, hey, make a bunch of sneakers or drones or, you know, landmines. They can do whatever they want. But that pales in comparison to the innovation and the drive of free men and women and capitalism.

Starting point is 00:42:22 And that's why we are going to win. And I agree 100%. Last thing on that is just that venture capitalists, and this is like one area where they're really bailing out the government. To be honest, like, without the VC community, we would be really far behind our enemies. For the venture capital community, this is exactly who should be doing it. Because it's risk capital, designed for rewards, and designed to build better products and services than exist. If the government were to award those people at MIT or, you know, whatever engineering school with just a grant, they would not move as fast than a company run by you or Palmer Lucky or Elon or, you know, pick your great entrepreneur.

Starting point is 00:43:06 because they wouldn't have the drive and the competitive nature. And that is what is missing from government allocation. And capital allocation is a competitive sport. It's one of the great underappreciated things. People like David Sachs are competing against Sequoia, competing against White Combinator, our firm. We're all competing to get deals. The founders are creating a marketplace for us to compete for deals.

Starting point is 00:43:35 for deals and pay the highest price, you know, and to keep the stock price growing, all that's part of great American capitalism, which at its core is competitive. It's very competitive. I'm a big competitor. Yeah. That's why you're able to do this in 18 months. If you were working at a school and you had self-selected into a university academia, the people giving out grants, you would not have.

Starting point is 00:44:05 have the urgency of doing it in 18 months. It's the fact that you were going to run out of money in month 24 that made it get delivered in 18. You need the pressure. Need it. I'm curious about the manufacturing element about this because we're talking a lot about American capital, American ingenuity, American defense. Is the supply chain entirely domestic?

Starting point is 00:44:23 And is that possible? And then also manufacturing, how domestic is that? Yeah, we're going to be doing this all in America. It's mostly currently in America. And the thing is, like right now, this appropriation. So I do a lot of the lobbying. I live in D.C. and I do a lot of, besides fundraising, I do a lot of the lobbying and Hill work. So we're working with Congress right now to decide where we're going to put some of these buildings. I mean, this is great, by the way, Alex, because you and I talk here all the time on this weekend startups about job destruction. Americans who are out there, you were thinking you were going to work in the marketing department or communications or whatever your degree is, maybe even a developer. You know, maybe you need to actually go to a factory and assemble something in the real world.

Starting point is 00:45:05 And that's going to create so many jobs in America that we're going to look at this anti-immigration kind of moment in time and think we're idiots. Now, nobody wants violent immigrants in the country. We need another 150 million people in this country in order to man the factories, to stamp these factories, to make these weapons, to make iPhones, to make cars, to make rockets here in America. We need to be on an America 500 program, 500 million citizens. Now, we want to do that correctly.

Starting point is 00:45:35 I don't know how you feel about immigration, but I do think you're going to have a hard time finding factory workers. Yeah, it's going to be tough. But I think it's a very exciting product and a lot of people want to come work on autonomous robot guns. So we'll see how we do. A factory building stuff like this make per hour.

Starting point is 00:45:50 Like, I know when they work in like a car factory, they make $30 an hour, something like that. I don't know. I think yearly salaries are somewhere between like $90 and $125. Yeah. Wow. That's a great job. That's a great job.

Starting point is 00:46:02 That's a great job. That's a great job. That's own a home job. Yeah, it is a really good job. Have one parent stay home job and raise kids job. Like, that is back to classic American. One parent gets to stay home. Another makes enough money to pay a $2,000 a $3,000 a month mortgage.

Starting point is 00:46:20 We're going to reindustrialize, and the VCs are helping make that happen. All right. Let's go. Let's go. I want to squeeze in one more question before we have to go, Stephen. We can keep you on all day. But how soon are you? you're going to put together a GoFund me that I can take some of my money to put into.

Starting point is 00:46:37 So that way we can send these to Ukraine. That you bring it up is I get like all these emails all the time because I went on Fox business the other like a couple months ago for this and uh everyone always emails like what's the ticker symbol when can we get in how do we do it? Yeah. I don't want to. No, he wants to fund it to send to send to Ukraine to stop Putin from invading countries. I want to donate.

Starting point is 00:47:00 I want to just buy Ukraine, both frogs. No, you're in. So I will, let's do that. My co-founder will love you for that. When the war, when Putin invaded, the first month is actually right when we sold to DoorDash, same month. And he donated a Bitcoin so that the Ukrainians could buy javelins. Nice. So like my, you're totally aligned with my CTO.

Starting point is 00:47:17 I'll have to connect you guys. Awesome. Because in continued success, you're putting, you're hiring four engineers. Where can they go? What's your email? Because there's a lot of engineers listening to this. Yeah. Working on nonsense at, um,

Starting point is 00:47:32 Facebook, trying to get people to, you know, increase the click-through rate on some scam ad by 0.01% when they could be actually doing something important with their lives, like working at your company. Yeah, they got to get them. Email Steve. Thanks, thanks again. Email Steve at Allen control systems.com. We'd love to have you. We're in Austin primarily, but you can be remote, but most people are in Austin.

Starting point is 00:47:54 Most people are in Austin. Yeah, we're South Congress, south of the river. Oh, yeah, great. That's where I'm in Driftwood. Yeah. Okay, yeah. I'm down there like twice a month. You have to come to the range.

Starting point is 00:48:04 You should come to the range and see the product. I would like to come to the range and see this. I'll do a live this week and start up from there. And then we'll get some Terry Black's beef ribs. You can come to Alex. You can't do the drone gun shoot. I will buy you a ticket. I'll buy my own ticket.

Starting point is 00:48:17 Who cares? I got you. Let's definitely do it. I think you'll really love to get out there. Okay. I want to do it. All right, Steve. Thanks so much for coming on.

Starting point is 00:48:23 Everybody go work for Steve. Or I should say four of you a month. Go work for Steve and protect this country and protect democracy. Captain America here. This Recon Stardoms five guests. He was great. I could talk to him for hours. Let's have Steve back on in one year.

Starting point is 00:48:40 You know, we have so many great guests that I was like, you know what? Let's try a two-for on this week and startups. So let's do a two-for. Let's do a two-for-tuesday. Coming at you here on Z-100, the morning show. Two-for-Tuesday, remember that on the radio? It's Wednesday, Jason. I know.

Starting point is 00:48:57 We all know that. Okay. Two-for- Tuesday in my mind. Okay. Would you remember two-for- Tuesday on rock radio? I do not. That might predate me. Two for Tuesdays were like, okay, coming at you, money for nothing from dire straits on a two for Tuesday.

Starting point is 00:49:12 And after that, the Sultons of Swing, yeah, coming at you. So my version of this was Mandatory Metallica on the local radio station. They would play three Metallica songs in a row at like 9 p.m. or something. I met the lead guitars from Incubis recently. Oh, really? Yeah. How was that? just friend of a friend and we talked about parenting

Starting point is 00:49:36 and he he's a cool dude like Mike Einzinger I met Mike Einzinger okay E-I-N-Z-I-G-E-R and we just went back and forth talking about rock and roll with our kids and like what bands we were

Starting point is 00:49:57 introducing them to so shout out to Mike Eindsinger This might be a new celebrity bromance I'm having. You know, I have a celebrity bromance once in a while where I meet a celebrity and then I bro out with them. And then you make them come to your poker games, then you take all the money. No, not money, but, you know, more like just a hang. But yeah, it's a little bit of that.

Starting point is 00:50:19 But interviews, Jason, let's have our next guest on because this is one that's actually very exciting from the launch perspective. We're going to talk to Rami Abiyhabib, the CEO and co-founder of a company called Quarry. Hey, there he is. Hey, Rami. How are you, sir? I'm good. Thank you for pronouncing my name, right?

Starting point is 00:50:37 Very excited to talk about the second half of American dynamism, B2B SASS. It's very exciting. No pressure, Rami. What are you doing in the world? What did you get done this week? Are you protecting democracy and humanity? In my own way. In my own way, I am producing jobs and I am protecting democracy.

Starting point is 00:50:54 All right. Absolutely. Well, welcome to the program. Tell us a little bit about what you're working on. and maybe even show us a little bit about what you're working. Yeah, sure. So we're to talk. I'm Robbie.

Starting point is 00:51:04 We're building Quirio. Quirio is the best way for teams to work with data. We really try to make sure that everyone can work with data at a way that matches their technical level. I think over the last couple decades, we've generally perceived data as a very transactional relationship. You have someone else. Bring you what you need. Most people can't access themselves. The literacy is quite low.

Starting point is 00:51:24 And year by year, data becomes more important, right? one of the only jobs that are still higher in demand than supply or data jobs, data scientists, data analysts, et cetera. So how do you make that new generation tool where those technical people feel like their needs are met and the business people feel like their needs are met without compromise, I think is the biggest thing. So can you show us, like maybe take the screen over here and show us the product, full disclosure. You went to our accelerator.

Starting point is 00:51:51 I'm unsure if you went to Founder University before that. I did, I did. You made me quit my job. And then after that, I built a company and then got into the accelerator and now we're here. Okay. So maybe before we even go into your product, let's explain that journey for a second because, you know, for people who don't know, I found myself as an early stage investor when Sequoia Capital asked me to be their first scout along with a gentleman named Sam Altman. He did Stripe. I did Uber.

Starting point is 00:52:20 And that got me started as an investor. And one of the things I found over the years, Alex, was I got up to 4 or 5,000 applications for funding a year, all in Brokout became a public phenomenon, crossed over into public mine share. And I went from 5,000 applications a year to 20,000 coming into our database. But the problem was a lot of those companies, Alex, there's a lot to go through. So now I have seven full-time researchers and analysts going through those, putting them in a database. Well, they're already in a database, but, you know, sorting through them and meeting with the top. top four or five thousand of the top 20,000. But what we found when we looked at the data, Alex, was half the companies were incredible teams that hadn't incorporated yet. But they had a

Starting point is 00:53:04 killer idea. And they had two or three co-founders who were technical product managers, just awesome. We had no way to engage them. We'd say, oh, let us know when you raise money, I guess. And then I was like, you know what, I'm going to come up with a solution. I came up a founding university, a 12-week course. People come to it. They don't need to be incorporated. We invite 250 teams. And then we'll just invest in whichever 10% to 20%, 25% to 50 companies.

Starting point is 00:53:30 We'll give them 25K or 125K as their first investor. Because I would like to be the first investor in another unicorn so that people could stop saying I was the third or fourth investor in Uber. And I can just have, and I think your company, remember, we that. Anyway, long story short, this has helped us engage with another thousand. companies a year and invest in another hundred companies a year because we have a pre-accelerator. That's the setting here. So how did you find out about Founder University?

Starting point is 00:53:59 Yeah, so funny enough, I found about it on X through U-actions. I was falling and I saw about Founding University. And so I applied at the time I was working at Amazon. I had a great idea. I was kind of like the personification of Curio as a person and I worked in a lot of technical and business roles. And me and my co-founder, who I met at UT Austin, actually, and lived in office. Austin for a few years. And I think Salt Lake in Wemberley is better than Terry Blacks, but I'll leave that for other people to judge. I go to the Salt Link frequency. It's about 10 minutes from here. Nice. Yeah. And yeah, I applied for a funny university and got in and ended up, you know, we had built an MVP. It worked really well. Got a lot of good value out of the, you know, program. And then we ended up getting investment from the pre-excelerator. And that was really kind of like that big push that made us incorporate. We both committed to career full time, quit our job.

Starting point is 00:54:49 jobs. My co-founder, he had worked an operator at a couple of exited startups, and it really just lined everything up for us and really gave us that initial push that we needed to keep going. So just to be clear, my thesis was, hey, if I'd build this thing, found a university, I might get people to quit their jobs at Amazon or Uber or wherever they are and start their company, and that first 25 or 125K check could be the thing that just, you know, helps them make the jump. And in this case, that was in fact what happened? Yes. Yes.

Starting point is 00:55:21 The plan works surprisingly. Yes. Literally warms my heart that we made a plan, we executed on it, and it happened. Which is very rare. That is really rare for a plan to work out, but yes. I love it when a plan comes together. If you're a child of the 80s, you remember the 18. I love it when a plan comes together.

Starting point is 00:55:37 I got to get my cigar out. So show us the product here. Enough with the, we've done the self-congratulatory victory lap. here, an idea worked, but let's see what your idea is. Let's make this about you now. Absolutely. I will show you the current version of Quirio. Maybe we can talk about how it's going to be changing. But yeah, so first and foremost, Curio is a BI tool, business intelligence platform is to help businesses analyze their data and query their data. So today we're going to be looking at some data for Dunder Mifflin. For those not familiar, that is a very famous TV show called The Office

Starting point is 00:56:11 so I can ask Curio, you know, who am I? And it should tell me. who I am. Of course, I'm going to be the best character on the show, which is obviously Michael Scott. He is hilarious. So yes, we are Michael Scott, regional manager of Scranton PA. That's fantastic. Pretty much like other classical BI tools, you have things like, you know, dashboards where you can track things that you're creating that refresh automatically that you can filter. What are the classic BI tools? What are the top two or three people would recognize as business intelligence tool. Looker,

Starting point is 00:56:44 Tableau, maybe SAP, if you're in a very old organization, those are kind of of like the more larger ones. And you have some new age ones like Thoughtspot potentially, depending on how modern your organization is. But I think Tableau and Looker would make most people's brain kind of light up.

Starting point is 00:56:59 Business intelligence started in 2000 or so, 2000 to 2010. Those companies started to emerge because of big data. Yeah, so it was like SAP really started the whole, you know, online chart creation where a data team was making these charts that they could send to their business counterparts for them to track things. That was really like the first big one. Second big wave came with Tableau. They let you actually filter them. Looker let you look at the tables

Starting point is 00:57:20 underneath them. Every five years, you got a bit more control on the business side. And this is where an entire career emerged for people who would be in big data. What are the job titles that were created because of this over the last two decades? It started off with like business analysts who would just analyze business data, data, data analysts would be a little more technical. You have the engineers who would actually move all the data around, the data engineers, and then you have the data scientists who forecast do deeper analysis that became more of machine learning when that came out. And then some ad hoc ones, you have like developer analytics and then some random kind of more niche ones.

Starting point is 00:58:01 Got it. Yeah. Okay, so that's great. So this entire industry has boomed over the last 20 years. Alex, to give context, that 10 years before that, we saw things like databases. and storage and the internet kind of get was the setup to all of this, right, Romney?

Starting point is 00:58:17 Like the fact that people had cheap storage, the internet, there was a lot more data recorded and then people said, well, what's in this data? And a classic example, that would be somebody like Amazon who has sales data, they have a lot of data. Too much data. In 1998 till today,

Starting point is 00:58:35 there are so many databases at Amazon. It is crazy. Yeah, you know, you can't fit things in an Excel show, right? You have over a million rows, you're getting hundreds of millions of new roles every hour, every day. So part of the setup is all this data exists. You have to go to somebody like a developer, get them off of building a product, ask them questions, and wait a week or two.

Starting point is 00:58:57 Then you forgot that you put the request in. And then they come back to say, hey, I did about five days of work. Here's your data. Yeah, only for them to not actually be in your day to day, not understand the nuance of your question. And then you have to tell them, okay, that's, that didn't pass a sniff check. I think you looked at this wrong, can you pull it this way instead?

Starting point is 00:59:13 Okay, so that's our setup where you were showing up. The average ticket success time at Amazon, fun fact, was about five weeks for a general data request. Yeah. That's like two years. Five weeks. Yeah,

Starting point is 00:59:24 that's an insane amount of time. It's insane. It's ridiculous. Romney, if I understand that, the point of query is that it brings in a lot of traditional BI tooling charts, dashboards and so forth,

Starting point is 00:59:35 but it also lets me query it in plain English so that way I don't have to go to a data analyst or data scientist, I, myself, the person in a business role, can do it myself. Yeah, so that's one part of it. I think, you know, I think it's really great that companies are trying to be more data literate.

Starting point is 00:59:50 I think it's great that we're now buying data products for companies. But, you know, do you look at Tableau? I think something like 95% of the licenses they sell are just for viewers. You can't even make anything. You can't even drag and drop columns to make charts. Wow. And so it's great that we have all these tools. We agree everyone should have access to them,

Starting point is 01:00:06 but most people can't actually use them. They can just kind of open them and look at something. And so, you know, we're trying to build that BI tool. We are building it where how do you make the data people's lives better? You know, not only are you going to save them time, how do you make their workflows actually better because you want to convince them with more than just you'll have a few less questions. And then how do we make those business people kind of flip that on the head? And instead of having only a transaction relationship waiting, that you get to do some more work yourself.

Starting point is 01:00:30 And, yeah, so, I mean, we can jump into a couple of questions, see how Curio would answer them. Okay, yeah. So we have this, we have some data here in Quirio. It's Quirio. dot AI, just if you want to go look at the website. And this is year one of your company, basically. Yes. And you've built your, what was an MVP when we met, now it's an actual product that people can buy. Yes, and customers hands and people have bots. Yes. Yes. Wow. Amazing. That's great to see that all happen within under 18 months. So what can people do with it?

Starting point is 01:01:00 Because this is so you don't have to wait five weeks, right? That was when you originally talked to me about it. It was like, hey, there's no reason for people. you know, 100% of people in our organization should be able to make these queries of the data sets and we want to make a tool

Starting point is 01:01:16 that takes out that telephone game, yeah? Yeah, we want to take a tool that takes out the telephone game. That's exactly it. And also a tool that will serve what the current needs are, but even better.

Starting point is 01:01:26 So let's start with the one we're generally talking about, which is you don't need to have this telephone game going back and forth in someone technical and getting into the context in the business team.

Starting point is 01:01:33 So yeah, we have some data, we have some account data, orders data, Dundermifflin sells paper products to a bunch of different companies. So we can ask a question like, you know, let's list the top 10 products by revenue and give us their quantities sold. Pretty basic question is we can ask it. Queryo can look through the database, understand how it all joins,

Starting point is 01:01:54 understand how it works. It'll tell you what is going to do in English, write the code for you right after. That's just a SQL query and you can see it if you're technical, you know, just so you can be sure that it's right. And then just like that, You get your answer really quick.

Starting point is 01:02:09 You have some classics up, whether you can sort the data, whatever it is, which is really nice. But you can do more than that, right? This is a query in question, like just asking, show me the X of these. It's kind of like clicking through some Excel sheets, but through a database that you can't see.

Starting point is 01:02:21 What if we had a question like, you know, are there any salespeople that have to offer some discounts to close a deal? You know, which account matters are offering the most discounts. Yeah, I'd like to know that myself. Yeah, so that's my bet is Michael Scott probably. He seems that the kind of guy who would offer a lot of discounts. So yeah, it's going to join the orders and account table, for instance, on account ID. So it has an idea of how all the tables relate, which is a really complex thing that a lot of data team struggle with sometimes.

Starting point is 01:02:46 And then it's going to calculate the average discount percentage. So just like that, we have a list of all of our account managers, the average discount percentage. It looks like Aaron Hannan is actually number one and Michael Scott is third. So, or actually he's tied for seconds. I like my guy Ryan at the bottom because Ryan just holds the line. Ryan really does hold the line. We don't discount. That's it.

Starting point is 01:03:04 No discounts. We charge a fair price. He holds the line. Yeah. And I guess what's really cool with these kind of products, you can think of AI as the best data scientist that ever existed, but it's the first day at your company. And so the biggest thing we have to solve is how do you get this digital kind of

Starting point is 01:03:21 context understanding of how your company and your data works so that it can write the correct queries, right? It's not looking at the rows of data. It's actually just writing the code instructions to pull data from a database. And we do that through something called our knowledge base, and it's basically where, and here it is, is where we kind of capture all the details of every column, how all the tables join.

Starting point is 01:03:39 Usually column names aren't very descriptive of what's actually in them. And I think maybe about like 15% of my time at Amazon when I was working with data was just trying to find it. And so this is great in that end. So, Rami, on that point, though, how hard is it to onboard a new customer into this, into queryo? Do they have to change how they upload data into the service? Or can it parse a variety of different, I don't know,

Starting point is 01:04:03 inbound sources from different companies that might have different approaches to storing their information. So to start off how hard it's funny, it's actually the more mature the company is, the easier it is. So if they don't have a data team, they're maybe only running with a CTO,

Starting point is 01:04:18 maybe some engineers like flexing, doing two roles, and then really have the time to set it up. That's where we can have a bit more of a manual approach, unscalable things when you're starting off. But data teams are quite used to building things called data models and data catalogs where they actually kind of document

Starting point is 01:04:33 what's in the data to make sure that it works. And we have our own system and how we define things so that they maximize the efficiency for the data teams and also the automated like AI process. There's nothing they need to change at all. So it's something they just, you know, like fill out in their own way. You can also already have kind of schemas of how the data works and how they already join that pre-exist.

Starting point is 01:04:56 We have an agent system that can also fill out to do a first pass where they can check. And we can go into more details. But yeah, it's inversely correlated with the maturity of the company. I appreciate that. And does that actually help your early customer acquisition? Are you dealing mostly with larger, more legacy clients that are therefore easier to onboard? Or are you selling more to startups today that have more of an issue getting set up?

Starting point is 01:05:18 That's a great question. Yeah. So, you know, the whole thing we're trying to build here, which is meeting everyone at the technical level, you can see everything I demo today is really, really good for those people that aren't technical and need to ask open-ended. And so that process is working really well. But it's a, it's a, you know, cash money too. Since we are built for those non-technical people, they're also the hardest to onboard.

Starting point is 01:05:37 And so the next phase of the company over the next six months is we're building out a lot more of these features that those data teams are going to love. And then that'll, you know, increase our ACVs and then also lower our onboarding times in our sales cycles, which will be great. Is the way to go to market here. I always like to think about go-to-market strategies in year one of a company, you know, as you get into year two, you've got to really be. thoughtful about who is your ideal customer profile? And I wonder if you've given thought to, hey, I use tools like Looker, I use Tablo, and so I'm going to add this to my utility belts and put it into the mix. Or do you think, hey, our ideal customer is somebody who Tablo and Looker are too complicated

Starting point is 01:06:22 for and they don't have it yet, but they do have underlying data. In other words, they're kind of a cleaner. They're not like data blind where they don't care about data. They're just running some small business. They're not data-driven organizations, but they're like that sort of middle, mid-sized company. How do you think about which one to go after? That's a real question. Well, I think I'll start off by saying that any company that needs the internet to run is probably going to analyze data.

Starting point is 01:06:52 Okay. You know, outside of like maybe like some niche e-commerce stores, for the most part, it's a fundamental part of most businesses. and they can't avoid it at some point. I think we've had customers ready to leave products like Looker and Metabase and Cabana for us, which mainly means that we are taking the market share. We don't think we're directly competitive. But the idea is not to be part of the belt, it's to give them a solution that actually fits their needs a lot better,

Starting point is 01:07:19 which those tools just don't. So it's more the second one, which is to replace those tools and give them functionality and actually can solve it two or three tools instead of just, replace one. See, I think this is like a very interesting exercise, Alex. It's one I like to do with founders is just have this dialogue. So what you're seeing here when I, when I ask Rami a question like this is I have my own thoughts on it. I like to hear the founders thoughts on it. And then I can ask another probing question. And this is what eventually results in a board of directors creating a plan. And then that plan is for the next year. And you,

Starting point is 01:07:58 execute against it. And then you put priorities and resources against a plan. And as Doug Leone told me, hope is not a plan. Plan is a plan. Make a plan. And so now when I look at these things, I have a really interesting, humble approach. I just listen to founders. And when I hear you talk, it was just like a bunch of things burned in my head. I realize there's some brilliance to the simplicity of your product, that even people who could figure out a complex legacy product, and a legacy product tends to be any product over 10, 20 years, any product that's over 10 years is now a legacy product because interfaces change, AI changes, there's so many different things to do it. And it reminded me of a discussion I was having with a friend of mine, when you get to

Starting point is 01:08:48 a certain scale house, Alex, and you've experienced this, I think, with your in-laws and some other folks, you mentioned, you will be told you have to install a home automation system like Questron or Savan. These things cost hundreds of thousands of dollars to implement, but on a $10 million home, you're like, well, it's one or two percent of the cost of home. I need something to manage this home. Nobody ever uses them. They're always broken. They require somebody for two or $300 an hour to come in and reprogram everything. And you get to have this moment where you have a weird remote control people come stay at your house they can't turn on the news or watch the next game or put on music every time i buy a new home i rip the shit right out else and you know what i do so no synaptle tv

Starting point is 01:09:34 yeah we're done you're that rumor that apple's releasing a like a control center for homes i predicted this two years ago we did something what should apple do next if they're not going to do project tight in the car i said the number one thing they should do is home automation because i use apple home it's garbage i use google home it's okay I love Google's drop cams, which became Nest. They screwed that whole thing up. But I just go with Sonos because it's elegantly simple and it's beautiful. And I go with Apple TV because, again, everybody knows how to use it.

Starting point is 01:10:09 And so I think there's something to simplicity and UX and design that you may have hit on that would even make the person who wants to put in the $250,000 savant system or Control 4 falls. in this bucket where you're not allowed to make changes to it. You need a technical person, blah, blah, blah. Screw it. Like, go with elegantly simple. There's something to it. I mean, you're looking at one of our first customers.

Starting point is 01:10:34 They were paying 60K a year for Looker, and only the CTO was second goal enough to use it. Oof. From a seam of 20. And it's ridiculous. And not just that, they paid for a separate tool. They were also paying for Hex because Looker doesn't support Python. So we haven't actually gotten into the non-AI-I-R part that we're improving.

Starting point is 01:10:51 There are some BI-te. tools like mode analytics, they got bought by Thoughtspot, but the legacy tools don't even support Python natively, which is not an AI thing. They just only do SQL. And data teams are not having to buy a Python tool like a Jupyter notebook, like host, and a SQL tool that has

Starting point is 01:11:06 lets them host dashboards like on Looker. And it's ridiculous. And they're paying six figures at sometimes seed, definitely at Series A, for like three people to have some good data. How much cheaper can you go compared to those kind of legacy price points? I'm curious how much of a

Starting point is 01:11:22 price war we might see as you guys take on these incumbents. I think the tools are decently complex to make where there's not going to be like a flood of these really kind of consolidating Python with like good AI agents with good SQL IDs. So right now our average ACV is about 8K, 8 to 10K. Our two newest customers were at the 15K. Ideally next year we can start reliably closing those 20, 25Ks.

Starting point is 01:11:49 I would be... A 50% off as a baseline, I think. is the disruptive pricing. Yeah, and it's great because if we can get up to their current ACVs, but offer, you know, you can also get rid of this other tool that has nothing to do with Looker. Yeah, now you're 90% discount. Yeah, now you're not such a great value problem. And we're going to get more value out of your business team.

Starting point is 01:12:06 They're going to be able to self-serve more. You know, you data team or CTO, you're going to have a lot more time on your hands. And the other Python works going to be quicker. There's just a lot of value to claim from so many different ends, which is really nice. All right, Rami, this has been great. I want to let people know three things. if you, like Rami, are working at a company and you have a great idea. I'm waiting for you, founder.

Starting point is 01:12:26 That university. Number two, if you have a MVP, you got a couple of beta customers, I'm waiting for you. Launch Accelerator, launch.co slash apply. Number three, once in a while, if I have a great company, we will share that company with other angel investors and help them get in early. And that's done through something called the syndicate. I was able to get the IP around the syndicate after I left Angel list. So I have the Syndicate.com, 11,000 angel investors. I think 4,000 have actually done a deal with us.

Starting point is 01:12:57 We've done 300 of these deals. So go to the syndicate.com and you might see companies that come out of our programs who are raising small amounts of money at reasonable evaluations to kind of get to the next step. If you're a high risk accredited investor, you can join the syndicate. There's a waiting period, like 30 days. And then you have to be accredited right now. maybe in the future that changes the syndicate.com,

Starting point is 01:13:20 go apply now. And worst case scenario, Alex, you read the deal memos and you say no to everything, but you get smarter. Robin, thank you so much

Starting point is 01:13:27 for including us in your journey. I appreciate it. And I'm looking forward to great things to come. Thank you. It was great talking to guys again. And for everyone listening,

Starting point is 01:13:39 I highly recommend all the different programs Jason has. I've done pretty much all of them. I am the spokesperson. You've checked off, I think, three or four boxes. so far, maybe you'll check off the fourth at some point soon.

Starting point is 01:13:50 All right, well done, Rami, and it's Quirio. .AI, spelled. Q Ue, R-IO, similar to query. Dot A.I. Quirio. dot AI. All right, we'll see you soon. Take care, Rami.

Starting point is 01:14:03 Thanks, guys. Bye. Alex, I think we've done enough show for today. We have, we're going to be back tomorrow with a lot of interesting topics. We're going to talk a little bit about exits and IPOs. I saw an incredible stat and have some great charts. that my friend Fawad was sharing with me on WhatsApp. So we got some great stats and charts for you tomorrow.

Starting point is 01:14:25 We'll have the Nvidia earnings tomorrow. Anything else on the docket tomorrow that we should tease right now? Wouldn't hate to go over Surface Titans S1 filing. There's a new Microsoft AI deal with Harper Collins. That's pretty interesting. I'm a Harper Collins author for this book right here, Angel. And I want to know about that deal because I haven't gotten a call from Hollis yet.

Starting point is 01:14:49 I don't know if the dollar amount will actually be very interesting to you, but the pitch is, and there's just, there's a bunch to get to tomorrow, guys. So if you do love Twist, we are live tomorrow at 12 Central,

Starting point is 01:15:00 yeah? Yeah, 1 p.m. Eastern, 10 a.m. Pacific, West Coast, or as we would say in New York,

Starting point is 01:15:06 the left coast. He's at Alex on Twitter. If you know who the main character is or a deep cut of an interesting story, you can always DM him. I assume DMs are open. My DMs are open and man, Are they messy?

Starting point is 01:15:18 Or you can just mention both of us, at Alex and at Jason. We'll see you all tomorrow on this week in startups. Bye bye. Bye.

This Week in Startups - TWiST News: Venture-Backed Defense Startups, Anti-Drone Guns, and Querio's AI Data Platform | E2048

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.