This Week in Startups - The Power of WEKA with Antonio Gracias and Liran Zvibel + Jam with JCal: Uptrends AI | E1986

Starting point is 00:00:00 They go between 5 and 12% of the fund. You know, this is just at 5. We would have bought more if it could have. The round wouldn't give us anymore. So, you know, I'm hoping after this podcast, he decides to sell us some more shares that this round was super oversubscribed. And so we were scratching and cloning everything we get. We've offered do some operations help, Rebecca, which I think while we were able to get the

Starting point is 00:00:16 allocation we got. But we were super impressed these guys and with what they've done. And you'll, you know, you hear about the customers, the logos, the growth. I mean, I think you'll feel the same way. And this is a strategy that some savvy managers are pursuing now, just if you have that when they're trying to get as much into it as possible because of the power of law. I see Brian Singerman doing it. I saw one time Sequoia deal with WhatsApp.

Starting point is 00:00:37 But I will say we invented it. You invented, yeah. And we've even taken a page out of that book with, you know, our fund instructions. So yeah, great artists copy. Thank you. What is it? Immature artist copy. Mature artist steal.

Starting point is 00:00:50 This week in startups is brought to you by Command Bar. Seamlessly integrate an AI-powered guide into your software, making navigation intuitive and interactive. Visit command bar.com slash twist to get a custom live demo. Dot Tech domains. Don't miss our Jam With JCal contest. To apply and get more details, go to jam withjal.com. Brought to you by dot tech domains.

Starting point is 00:01:15 And Lemon.io. Higher pre-vetted remote developers. Get 15% off your first four weeks of developer time at lemon. At lemon.com slash twist. All right, everybody. Welcome back to this week. in startups. I'm your host, Jason Kalakanis. And I'm really excited today because a friend of the pod, Antonio Gracios, is here. He is a tremendous investor, one of the greatest of this generation from

Starting point is 00:01:38 Ballar Capital. You know some of the amazing investments he's made over the years from SpaceX, Tesla, and now actually an investor in Athena, an investment we just did together. And, you know, we saw along the newswire press release about this new investment you're doing in Weka. And so I thought, Gosh, this seems like a very important company. And so I asked you and the CEO Antonio to join us. Laurent Zubel is the CEO of Weka. It's pronounced Weka. And there's some background to the name, yeah, for people to remember Loran.

Starting point is 00:02:11 Welcome to the program. Thanks for having me. So Weca is pronounced as any Greek unit. So you probably know how to pronounce Mega, Terra, Xa. So you're just doing your Weka. Weca is 10 to the part of 30. and if Xe is a thousand peta, which is probably the biggest number people really can interact with, WECA is a trillion XA. So it's a huge, huge, huge number.

Starting point is 00:02:34 And you know, Bill Gates at once said we'd never end up getting more than 640 kilobytes in a computer. Why would anyone need that? I'm not going to predict whether we will or will not need a WACA, but hopefully we're all going to leave to the point that someone's going to need that. And so as we start here, Antonio, we're always interested when a capital allocator who likes to tackle hard problems comes into a company and writes a big check. So maybe you could tell us a little bit about what you discovered here with Weka and what the investment thesis is. Thanks, Jason. And thank you for having us. We're really grateful to be here. It's always enjoyable to see you and talk about our companies. So, you know, I think first and foremost, we're spending a lot of time in artist intelligence. As you know, we invest in Deep Mine back in 2010 and I was on the board of Tesla during the being of the eye vision systems there. We thought about that a lot, and we have dozens of investments on our intelligence. They're focused in two areas. One is infrastructure.

Starting point is 00:03:29 The other is what we call verticalized applications. These are applications that have very good criteria data, and then a reinforcement is very tight. Weka is in the first category infrastructure. And what they're doing is making the data center much more productive. So the average GPU is utilized about 30% of the time. This is like a close guard and secret, but that's what we think. Weka makes that a lot better. and they do it in a very, very interesting way.

Starting point is 00:03:53 I'm going to let Loran tell you about the product. But I will tell you when I, a funny story of Loran, when I first met him, and we actually went the second, the second time met him, we went through the product. I asked him to explain the math to me out of the same work. It's kind of amazing how it moves. It makes the back end of the GPU, memory behind the Gron is much more productive. He went into the math. And he started diving into the math, deeply on a whiteboard, like a, like a, you know,

Starting point is 00:04:13 kind of a, you know, kind of a professor kind of thing. And I'm watching this. I'm asking some questions, and I realized that I actually remembered my math. I knew what he was talking about. So I was very impressed, but he's a deep product experts. The first thing is the product's amazing. It fits inside of our investment thesis for the infrastructure and the data center. The second is that, you know, our strategy, as you know, is to write kind of a smaller check.

Starting point is 00:04:35 You know, a company, bigger check, bigger check, bigger check. We wrote a small check in the WECA followed them. They beat every number they ever gave us and then exceeded substantially. Product market fit was incredible. And then I got to know Laron because, you know, we got no companies. I was impressed with Lehronda's team and how well they execute. I mean, they are really product market experts. They're product experts and they built something before the market knew it needed it.

Starting point is 00:04:58 And then the market came to them and the company just exploded. That's what we did. And this strategy to feel or bet, get to know the company and then place a second or third bet, it's worked before. Yeah, it's worked very well. I mean, you know, our portfolios, typically our biggest winners are large investments, very different than many managers to kind of spray out. lots of positions. We like very contrary positions. We'll write, you know, lots of small checks, but our big checks, and this is a hundred million dollar investment in our fund. That's a large

Starting point is 00:05:27 position for us. They are typically our biggest winners. And those tend to be 10% of the fund, 5% of the fund size 15 when you make a gigantic, you know. They go between 5 and at 5 and 12% of the fund. You know, this is just at 5. And we would have bought more if it could have. The wrong wouldn't give us anymore. So, you know, I'm hoping for this podcast. He decides to sell us some more shares that this round was super or subscribed. And so we were, you're a scratchy and cloning everything we get. And, you know, we've offered do some operations help for WECO, which I think why we were able to the allocation we got. But we were super impressed these guys. And with what they've done and you'll, you know, you hear about the customers,

Starting point is 00:06:00 the logos, the growth. I mean, I think you'll feel the same way. And this is a strategy that some savvy managers are pursuing now. Just if you, if you have that winner, trying to get as much into it as possible because of the power law. So, uh, but I will say we invented it. You invented, yeah. I see Brian Singerman doing it. And, uh, I see Brian Singerman doing it. And, uh, saw one time Sequoia deal with WhatsApp. So I think collectively people are starting to become aware of this, Antonio. And you, I do give you some credit there for pioneering it a decade ago. And we've even taken a page out of that book with our fund instruction. So yeah, great artists copy. Thank you. What is it? Great artist copy. Some of Steel. Something like that.

Starting point is 00:06:40 Immature artist copy. Mature artist steel is like the strategy here for us and for maybe some of our contemporaries. All right. So let's get into the product. The company's exist. for a decade. I know you've been moving large data sets around. So explain to us, Leon, what happened in the history of the company where you realized, hey, managing large datasets, moving them around, and then all of a sudden these H-100s and AI sort of hit a tipping point. I think we'd all agree about two years ago. Maybe you could tell us a little bit about the strategy of the company here and getting that extra 70% out of every H-100 for people who don't know. People call it a chip. It's more like a platform made by Nvidia, and Zuckerberg just announced they did their latest LLM on 16,000 of them. So to get an extra 70% from those 16,000 is they're not cheap, right? 30, 40,000 dollars each, I believe, or on average. So yeah, tell us about the product and how you got here.

Starting point is 00:07:34 We've been doing data management. We've been doing storage for a long, long, long time. Work as their second company, the previous one called X-AV storage and IBM acquired us. So, Anyone you know that's been buying high-end IBM storage in the last 15 or so years in New York, previous company, we've learned a lot about the enterprise, about big customers and about why people buy storage while being at IBM. We left IBM at 11. At that point, we basically took an oath and said, hey, never again. This market just doesn't make any sense. And it didn't make sense for two main reasons.

Starting point is 00:08:12 One, the delivery mechanism for this product is flawed, while most of the engineers are software engineers, and when I'm saying most is like 99%, not 51%, the delivery mostly comes in some proprietary box. It makes it hard on premises, it makes it impossible to run the product well on the public cloud. The second problem is that the market is fragmented like crazy. So it's a $150 billion market, but there are hundreds of products that actually sell. And when you're looking at from the large enterprises, the big corporations, they're using anything between 10 products if they're really lucky to maybe 40 to 50 products across hundreds and thousands of silos. So it's hard for the customers. It's incredibly difficult for the vendors. So that's the reason we said we wanted out. Then back in 2013, when we started three huge, huge, huge changes in what's possible happen, containers happen. And they led to completely different thinking about software, microservices that led you take a big monolithic system,

Starting point is 00:09:19 chop it down to tiny little pieces and run them on many, many, many machines. So microservices changed how you can think about infrastructure. NVME happened, which allowed you to take flash, connected directly to the CPU, and control it very effectively in tiny little fragments. And then the third change, which was the most important one, is network finally caught up to the speed of the servers. So throughout history, the network may have been 100 times lowers if you're looking at the 80s or 90s, maybe 10 times lowers if you look at the early 2Ks. In 2013, network is caught up.

Starting point is 00:10:03 And by the way, by now, the network is twice as fast as the servers can create frames at times. 100 gig, but the network switches them 800 gig. So this is incredible. Founders, I know a lot of you listening to this podcast. You build software for a living, right? We love doing it, but we all know it's hard. You know the pain of trying to onboard new users and get them up to speak quickly. Worse, most chatbots and guides built for the task are annoying as heck. Users tune them out because we all hate random pop-ups. Thankfully, there's one company that uses generative AI to help users onboard without annoying them and it's called Command Bar. It has a chatbot that gives personalized responses to user questions

Starting point is 00:10:44 instead of a basic Q&A. And it shows users around your product like a live guide, cursor included. Even more, Command Bar can detect when a user needs a nudge, giving them a product hint or special offer to close that important sale. Command Bar is used by Unicorns you know, HashyCorp, Gusto, Sixth Sense, Angelist, and others. So here's a simple call to action. You've got to check out this product, integrate an AI power guide into your software so your customers can navigate your product intuitively and quickly. Visit command bar.com slash twist to get a custom live demo. All of our competitors, the reason there are so many products out there, they are basically compensating for slow network by optimizing for locality. They're creating a storage array.

Starting point is 00:11:33 I'm sure you've heard about the storage arrays before. As a storage array, you have a bunch of controllers. Each controller owns a portion of the namespace, and you're picking whatever storage array you want to buy, but how the application basically spreads their data across the day, across whatever, and you want to break into controllers in a uniform way. The problem is, once you've picked that,

Starting point is 00:11:57 your solution now becomes a victim to the clients, because if all the clients would approach a single controller, it's the bottleneck. So you may have a hundred controller system, you're getting 1%. On the cloud, it's even worse. Let's say a 1% no sits on an instance that has noisy neighbors. You're not even getting your 1%. What we have done between microservices and VME and fast networking,

Starting point is 00:12:26 we've started as a clean sleigh, and we have created a system. It took us years, by the way, about eight years to get it to really run well and then a few more ways to matured, where what we're doing, we're looking at what the server send, the client send every split second, and we run the perfect load balancing of IO throughout all the NVME devices

Starting point is 00:12:51 and metadata throughout all of the CPUs, and we're showing that even on cloud with noise in ebors, we're getting perfect linear scalability. So if you give us twice the big wacker, you're getting twice the results, And that's the reason, since what we're controlling is latency, what's keeping the GPU servers start for data is high latency.

Starting point is 00:13:15 Since we're controlling the one thing that actually matter, or the only product that optimizes for the one thing that actually matters, we can get to the point that it doesn't matter how many GPUs you have. What is their access pattern? We can get them filled up to the 90th percent. So this becomes an acute problem when you're trying to solve major problems in the world, maybe self-driving or, you know, running a language model. People really didn't face this problem all that often previously, yeah? No, they have not. So now comes the big question of our product market fee and how do you take it to market. So we predicted customers would need bigger scale, would need bigger performance. We started these things through simplicity saying, hey, you don't need to buy these 50 different products to run all of your workloads because we're just going to run them well. Had we had to go through that phase, it would have been a much longer journey.

Starting point is 00:14:21 What are the biggest workloads today if you were looking at like categories? I mentioned self-driving. I mentioned language models. Those seem like super obvious ones. Am I correct that those are in the top three or four? For sure. So we're now seeing the explosion of GPUs. You've mentioned the H-100s.

Starting point is 00:14:38 Previously there was the A's and the P's and the Vs and the B. The bees are coming in the next. And they're used a lot in AI. we are showing that we can get AI projects running on, if you're looking at the cloud, we have stability AI running us. When they have switched to us on the AWS cloud, we're able to get their storage bill down from 5 million to 1 million, so five times cheaper. And as Antonio said, before they switch to us, their GPUs are running at 30%.

Starting point is 00:15:10 The biggest problem was that they needed to get more job down out of the existing GPUs that they had. they couldn't get anymore because there is a supply chain crunch. After the switch to WACA, we got their utilization up to the high 90s, so over three times more output, with five times cheaper, so 15 times better. So for AI, we're bringing big, big help. We have big car makers that are using us exclusively for many years, and we got their time to Epoch from a couple of weeks to four hours.

Starting point is 00:15:44 Explain time to epoch for people who haven't heard that term before. So time to epoch basically means, hey, we're going through another round of training, and you need to go through, hey, I'm ingesting the new data or it's already there. You're running a bunch of ETL. ETL is not sexy enough. He cannot charge for it. So now it's called MLOPS or LLM Ops, but the ETL portion of it basically takes your tons of tiny files and prepping them for the next job.

Starting point is 00:16:15 So you're reading and rewriting them in the good format. Now you're retraining one form of your data. So you're picking a subset of the data. You're sending it to tons of GPUs to retrain. So you need to copy the data into the GPUs. These operation takes a long, long time. And it's really great that now you can parallelize these on so many GPUs. While you're doing that, you actually need to save checkpoints.

Starting point is 00:16:46 So you're writing a lot because if it fails, you actually want to go back and you don't want to miss so much time. And at that point, at some point you're done with the next retrain. You have now optimized for another portion and you would want to go and run regression test and see, hey, I've optimized the model to slightly better if it's a car, let's say, junctions with always stop when it's raining and there is a crowd that casts a shadow over the cycles. And now you want to see, hey, with these crowds casting shadows, did we get anything else worse? So you now are running a huge amount of regression. You basically would drive your cars through all of the rest of your data.

Starting point is 00:17:34 So you want to see that this has worked well. And finally, this is one circle. And now you have a slightly better model, and you start all over because... It's crazy, Antonio, to think, like, we actually went full circle in the history of computing from... There's way too much compute, way too much storage, tons of bandwidth, and there's no application for it. And, you know, it's just YouTube is chugging along. Facebook is chugging along. Massive amounts of data, your iFoto, going to 4K, storing it all, sorting it all.

Starting point is 00:18:08 And just nobody seemed to have a use for us. And now here we are, Antonio, with a moment in time when we have supply chain, you can't get enough GPU, CPU power, you can't get enough storage, you can't get enough bandwidth to move it around. And we're doing technology runs like it's pro-systems in the 70s or something where you're time sharing and you have to wait to release your product and run another job and coming out with versions of it. I mean, is this going to last?

Starting point is 00:18:36 or do you think all of this incredible effort to make better chips, better storage, better operating systems and processes will result in us kind of catching up to the jobs, as it were? It's funny, when I was 12 years old, I went to computer camp at Michigan State University, and I had a hand in punch cards. That's how old I am. And you come back overnight after the run. If you made a mistake on it, it just didn't run, right?

Starting point is 00:18:58 And you had to go back and like, you know. Get in line. Yeah, exactly. You write the code up by hand and didn't use the binary bunch code. This is like air traffic, or actually traffic control system. for the for the GPUs right so filling up the CPUs are always full utilize GPUs and I think the answer to your question is I think this is always a problem because the technology is moving um and the software side the LLM side that the model technology is moving faster than the compute so absent a large

Starting point is 00:19:24 breakthrough like a quantum computer or something which you know it's always like it's always one year away or two years away right it's perpetual it's almost ready right is as all you run silicon I think you're always going to have this problem because the model seems to be moving faster than the ability to create compute. Now, it doesn't mean you won't go through some boom and bus cycles. You know, you and I were both on the internet, right? So I was building infrastructure back then, the connector business, right? So overcapacity, I blew up.

Starting point is 00:19:48 Then we were short capacity, went back up again. And you'll continue to have these kinds of cycles. But the artist intelligence, I think it's kind of a super, super cycle in the sense that it is much bigger than the internet. And the need for training data is huge. It's global. And it's not going to stop. We're just at the very beginning of this.

Starting point is 00:20:05 So I think it's a, you know, it's a long-term trend that will have a sign wave with a very steep upward slope. Yeah, if we were to compare it to, say, building out bandwidth, that was a 20-year story. And, yeah, we had a boom-bust cycle as WorldCom in all these places where they overbuilt the fiber and Google bought it for pennies on the dollar. People don't remember this. It was the early 2000s. Yeah, they went for it and laid so much fiber. And that actually became, I think, in some ways, the precursor for things like Netflix and YouTube. and your photo library not caring about bandwidth anymore.

Starting point is 00:20:40 I mean, it used to be putting any viral video on the internet would result in a $10,000 or $100,000 bandwidth upcharge from your provider. And then bandwidth suddenly became free and unlimited. And the storage, that was another 20, 30 year story. So if we were to parallel to those, this could be a decade or two decade chase to get more GPA. Yeah, at least. And with the added difference that the, in that case, the end use, which is sort of like, moving packets on the internet is kind of a fixed use and it didn't change very much.

Starting point is 00:21:11 In this case, the end use is changing a lot. So the applications of our intelligence are going to exploding, right? So right now we're in a very early stages of like it's an app on your phone. It's GROC and XAI, right? It's not a robot yet with washing your dishes. So imagine the amount of data required. Just the things we can imagine, forget things we can't imagine. Forget things we can't imagine what's going to happen.

Starting point is 00:21:32 But going from, you know, the GROC app on your phone to the robot washing your dishes is going to take a lot of compute. We don't have it today. It doesn't exist. And it's going to be building it is an enormous lift. And anything that makes the dentist and more productive is going to be very viable. Founders, these jam sessions with JCal we've been doing are huge success so far. We've seen so many great submissions. We picked to winner so far and we interviewed them on the show.

Starting point is 00:22:00 and we're still looking for three more amazing founders. Here's all you need. If you want to come on this program this week in startups and Jam With J-Cal, where you pitch me your company and I give you a bunch of feedback. Just two qualifiers. We want you to have less than $2 million in funding, so we want it to be a new startup, you know, one that needs help. And we want you to have an awesome dot-tech domain name.

Starting point is 00:22:19 So head to Jam withjKal.com. Jam, J-A-M-W-J-CAL. It's so much fun. The winners come on the podcast. I have to tell you, this is the greatest editorial segment on the pod for me. I love doing it. Because you pitch me what you're working on. I give you some unfiltered feedback.

Starting point is 00:22:33 We do it in real time. We ping pong back and forth. We pickleball. I ask you a question. You give me a response. You ask me another question. And you know what? We both get smarter and we understand your vision for your startup.

Starting point is 00:22:42 And hey, it never hurts to get your company on the number one startup podcast in the world. And we're working with our friends at dot tech because it's super cool to have a dot tech. Rabbit.com. Dattec. Rabbit.com. Everybody's using them. Heck, I use it for founder Fridays. Dot tech.

Starting point is 00:22:56 So here's your call to action. Tell me about your awesome. dot tech domain and startup and apply for the jam session with JCal contest today at jam with jkow.com. We're picking those last three winners soon. And so take us through some of the logos, Lauren, that are doing the most interesting things. When you look at your lighthouse customers, people who are doing the biggest, most ambitious things that you can talk about, maybe you can give us a little highlight, either your big lighthouse customers or the type of customer or the type of job, if you want to abstract it a bit.

Starting point is 00:23:27 Yeah. So I've already mentioned some huge AI project between deep learning, machine learning, the generative AIs. I mentioned stability. We have mid-jurney and a bunch of the other, very, very exciting forefront of AI. But, you know, GPUs are really useful for other means as well. So if you've been to Las Vegas, you've seen the sphere and you went to the U-2 concert. No, you two basically were the first show on the sphere when they got the residency. It was June.

Starting point is 00:24:03 They had to start showing up and performing in September. Their initial plan was to use their own infrastructure that owned tons of GPUs, but no previous way of thinking about all-flesh arrays. They started re-rendering their think. It would have taken them six months. At that rate, obviously they didn't have the six months from June to, September, they've switched to us, the switch to WACA, from the moment they got it on the floor to the end of the first render, it took them under four weeks.

Starting point is 00:24:38 So six times faster. But then what they've realized, hey, if you're taking a HD square thing that makes sense in a stadium, you're porting it to the sphere, which this magnificent screen, it just doesn't look good. So the fact that we were six times faster, the fact that they got the first render and they still had two more months actually allowed their artists to come in and reiterate and change and reiterate until these magnificent experience happened. I don't know if you've been there, but at some point, yeah, I've had some friends who went to see the dead and the YouTube stuff, yeah. Yeah, so the dead already dead and calls used us and they have these experience where like the whole thing closes as you. on you like a box or you're out there on the desert and now that experience is like

Starting point is 00:25:29 magnificent it's so i bet there's a several times all inspiring it's like being a medieval european and walking to a cathedral right that's how it feels but i want to say this it wouldn't exist without weka it wouldn't work and um it was so important that the ron got invited to the opening night because he his technology enabled it to work it wouldn't work otherwise and you know the dueling family made a bunch of bets here uh of technology that didn't exist before they built the sphere. And one of them was this back-end GPU problem that went to solve. Yeah, if you think about it, a 4K movies, probably 50 gigabit, if you were to download it.

Starting point is 00:26:04 And I don't know if those are 8K or 4K, I'm assuming 4K is enough. But how many screens are there? Like, if you were to actually make it into a television screen, like is it 100,000 screens, 10,000 screens at some large number, yeah? Huge. I think they said like 150,000 screens. Okay. Yeah. Every second of the show, it's over 400 gigabytes.

Starting point is 00:26:25 And they did some really, really impressive stuff. If you go there, you look at the YouTube videos. You can see Bono and the Edge singing inside of bubbles. And if you're watching it, there is no lag. So there's not even a single frame of lag. So it all going through the wacker, being GPU processed, through the wika, projected, all in 120 frames. where no notice spill like.

Starting point is 00:26:53 Yeah, I was about to say, I was going to say 30 or 60 frames, but if you're doing 120 frames, it's obviously four times as much as that. Now you're talking about just an impossible task. And they're doing some things in real time, like you're saying, they're stitching in the images of people, which is just unbelievable. And if you think about Pixar, which now has like, I think, maybe half of the all-time highest-grossing films, their biggest problem when Steve Jobs was starting that company, people don't know this, was just waiting for the render of 12. story and it would, you know, they would make story changes and they'd have to wait 10 days to have it output. And that was, you know, according to Ed Catmo, their biggest challenge was actually the rendering of it. I mean, what is it? What would it take a Pixar film to render today? A full-length film, do you think, using your platform and everything? We're seeing a lot of these

Starting point is 00:27:38 things go through and it's unlikely that you're watching something enticing late at night that didn't go at some stage through a waka. So a lot of the big ones are easikas. But just, just, just benchmarking it to like make these films now to actually render a film what do you think it takes now to render so they know a James Cameron what was James Cameron one that he did that was so they have many many dozens of thousands of CPUs they're now switching to GPUs they have huge networks so when you're looking at these impressive AI infrastructure no AI is all the rage it doesn't matter if you're running AI if you're creating a new movie, that the data center basically runs of the triangle of compute, network, and data.

Starting point is 00:28:26 And now that everything looks so good because compute is four or five orders of magnet faster, the network is four or five orders of magnet faster. And all of these folks that when they're taking compute the network to the extremes, they also need to do it with the data. And this is where WECA comes in. Going back to what Antonio said earlier, hey, quantum computing or neural networks. I went to school in the 90s back the newer networks and quantum computing both failed too far up. What has happened with GPUs is basically we won over Moore's Low.

Starting point is 00:29:00 So we couldn't create a single chip that has twice the transistors every 18 months, but with advances in networking, we can now create bigger and bigger and bigger computers that are disaggregated and scale. And it took us another 15. in years, but this is what we're doing. So the world stopped scaling up, we're scaling out. Which requires that bandwidth and storage and the sharing of the jobs. You're basically chunking all that data and then managing it. And then in between those GPUs, I mean, we're going to be seeing optics between them soon, yeah? In some cases, it's already out there. And so that's how large self-driving car data sets or, you know, protein folding is getting

Starting point is 00:29:45 moved around. Let's end on this, Antonio. When you're looking at, at and just sort of zooming back, okay, you got to manage the data loads. You got vertical applications here. There's another big piece. People are starting to talk about energy and data centers. These data centers need to be at a different scale, Antonio, than what we traditionally thought of and they need a different energy footprint. Maybe you could speak to what your team and you are looking at there in terms of opportunity for capital investment, opportunity for, you know, capital allocation. I'm just generally what these footprints will look like because it's different, yeah?

Starting point is 00:30:20 Yeah, look, we're going to, you know, 100 megawatt and gatewatt scale plants. And as you know, around the world, everyone's thinking of the problem, you know, the NGAs Supreme Court in the Middle East talking about this. And our friends in the UAE and Abu Dhabi built a four gigawatt nuclear power plant to power data centers here in the U.S. You know, we have a couple of companies, and I won't name them all, that are working on renewable energy solutions for the data centers. And I think this is very interesting.

Starting point is 00:30:48 You know, we have a power problem in the U.S. We were not going to have power to make this work in America. And so we have to work both on generation, creating more power, and actually getting it to get the grid stable. So we can both have power where we need and the power to use and have it where we need it when we need it. Now, WECA is part in this, is if the data center gets a lot more productive, you need less data centers.

Starting point is 00:31:11 Right. Three times more productive means two-thirds less hardware. Exactly. So basically the return on capital, you know, our investment models about RIC, right? So the return capital data start goes up a lot if you are using your GPU more. And very simply put, so if you invest in things like WECA that make the data set more productive,

Starting point is 00:31:31 you're also making the energy footprint much lower because you just need less data centers. How are people solving the problem today? because, you know, creating energy and nuclear power plants, even in China, you're talking about five, six, seven years. Maybe they could do it in three or four, but, you know, it seems like it's five, six, seven years. So we're not going to have nuclear power in time. So what's going to happen if you have, you know, take out the crystal ball here? Are we going to have, you know, data centers being moved to where nuclear reactors are putting them next to the existing ones and upgrading them? How is this going to get solved? I think there's two kinds of, there's two kinds of companies in the world. They're those that go slowly and kind of do things straight by the rules and book, and they will take a long time to find power and get it right. And then those that move fast and figure it out.

Starting point is 00:32:18 And we have a couple of companies in our portfolio that move fast and figure it out. One of them is doing it in a very, I'd say, creative and unique way. The other has been executing a strategy for some time of using a strand of energy to build data centers. So we have a company called Caruso that's building a data center now next to a traffic. to wind farm in Texas that's built because of subsidies can't, all the power can be used. And they're putting their data center kind of right and actually use that power and then using battery backup as well. And I have backup behind that for co-gen. So, but there's, I think then, I think this, we're underestimating in America, how much energy we have around in hydro,

Starting point is 00:32:56 in wind that actually isn't being probably utilized. It's feeding old industrial assets can be repurposed, which we're seeing in the smartest players. And this is, I think, one of the great things about this country is you have such a fluid capital allocation strategy. Sometimes people feel overbuilt. We talked about fiber earlier. So somebody gets a subsidy overbuilt solar, wind, whatever, hydro. And then that energy is going to find a customer eventually. And it's actually worth it to ship the GPUs to the energy as opposed to building the energy where the GPUs or where you want the GPU is correct? Yeah, yeah, it's correct. Because you can get that work built. I think this is a period where radar entrepreneurs will win

Starting point is 00:33:34 because they'll free this problem out, right? This is a serious problem, and you're going to see a shakeout. We're the best, smartest, scrappiest, you know, most commercial people if you're how to make work, and the other kind of slower-moving behemists, maybe don't.

Starting point is 00:33:49 Yeah. All right, listen, this has been an amazing dive into an awesome investment. Congratulations, Antonio on the team at Valor for finding another breakout company. And, uh, Lauren, thanks for the hard work. this infrastructure is obviously going to solve some great problems from humanity beyond being entertained at the sphere.

Starting point is 00:34:08 At a dead show, you know, this is the kind of stuff that solves self-driving, maybe drug discovery solves cancer, or maybe even tells us how the universe was created. So in advance of those amazing questions being answered, I thank you all for coming on the program. You're hiring, Lauren, I know, with all this money you're raised, you're adding 100 people or so. I read, how can people find out more and what top two or three positions? Do you need to fail acutely? Because we've got a big audience. Here are people looking to work at great companies. So you can go to our jobs on the WACA website. So we publish all of them.

Starting point is 00:34:42 Some of them also go on our LinkedIn. We are hiring tons and tons positions in engineering. We have engineering in the Bay Area. We have engineering in Tel Aviv with engineering in Bangalore, anything from infrastructure engineers to even kernel developers to front end to. to cloud. So if you're a good software engineers, we're looking for you. And the flip side, we're also hiring tons of very hungry go-to-market people all around sales, sales engineering, marketing, the field folks. So basically throughout the whole company.

Starting point is 00:35:21 Okay. So if you can make it or you can sell it, a seat for you, you know, I always tell people, if you got a chance to get on a rocket ship, take the seat and figure it out later. We'll see. you all next time on this week in startups. Bye-bye. Right now, startups have to do more with less. We all know that. And founders have to be smart with how they deploy capital. Investors are very tuned in to being capital efficient. So if you need great tech talent, but you don't have the time to interview dozens and dozens and dozens of candidates, you need to check out lemon.io. They have thousands of on-demand developers to choose from, and these devs are vetted and their experience, and most of all, they're results-oriented.

Starting point is 00:36:03 They're going to get you the result you're looking for. They're not going to leave you hanging. And guess what? They charge competitive rates. Great developers can be incredibly hard to find. We all know that. And when you do find them, it can be hard to integrate them into your team, but lemon.io will handle all of that for you.

Starting point is 00:36:19 Startups choose lemon.i. Because they only offer handpicked developers with three or more years of experience and strong portfolios. In fact, only 1% of candidates who apply, get in. And if something ever goes wrong, Lemon.io will get you a replacement ASAP. A couple of launch founders have worked with Lemon.com. And they've had great experiences. So here's your call to action. Go to lemon.com slash twist to find your perfect developer or tech team in 48 hours or less. And Twist listeners get 15% off their first four weeks. Stop burning money.

Starting point is 00:36:52 Higher developer smarter. Visit lemon.io slash twist. All right, everybody. It's time for another jam session. with J-Cal. This is a very simple project that I came up with. This is my invention. No, it's not actually. You know who's an invention it is? Travis from Uber used to do something called the jam sesh. And it was just like a couple of founders getting together, they hang out, you know, pop up on a couple of coal ones and talk about business and jam out on ideas. Man, it was some of the best times they ever had and we're bringing it back to this week in startups. And we have got a partner on this program, the partner is dot tech domain names, uh, and they came up with a simple idea.

Starting point is 00:37:33 Hey, listen, if you got under two million in funding and you got a dot tech domain name, which is a really cool domain name to have, uh, you get to come on the program. If you've got a great idea and a great company. And so today, we're going to hear from Ramsey Schaefer and he is the CEO and co-founder of Uptrends AI. And they are Uptrends dashaI.Tec. Ramsey, welcome to the show. Thanks for having me, JCal. Excited to jam.

Starting point is 00:37:57 Okay. Let's jam out. Why don't you start just telling us for two minutes about your company, run us through, it shows the product, whatever you want to do. And then tell me, what's the most challenging part of your business? Three, two, go. Sweet. Okay, I've got some slides.

Starting point is 00:38:11 Love to just get your raw feedback on them. And then I've got some questions that we can get into. Feedback on the deck. I get that a lot. Feedback on the deck. You're going to use this deck to raise money or you just want to explain the product? To raise money. Yeah, intro call.

Starting point is 00:38:23 Got it. Okay, good. So the audience for this is seed funds. I assume you're a C-Stage startup. So great, three, two, go. All right, I'm Ramsey. I'm the founder of Uptrends. We help financial advisors stay ahead of the news.

Starting point is 00:38:37 So this is Zach. He's an independent financial advisor, and a few times each day he'll get an email from a client asking him something like, you know, John Deere is up 5% today, why? Or I saw Nike fell 20% last week. What's going on? So he'll go to Google.

Starting point is 00:38:53 He'll go to Twitter. He'll go to Morningstar for some headlines, but more often than not, he's left scrambling to get back with a solid answer. Now imagine, multiply this by dozens of clients, hundreds of stocks, thousands of daily news events. And you can see how this becomes, you know, a huge time-consuming part of Zach's week. And frankly, it's holding them back from being a better advisor with more clients. So we're introducing Uptrends, the AI assistant automating the news cycle for investment advisors. Uptrends monitors, thousands of new sites, filings, and financial data sources

Starting point is 00:39:23 to detect, summarize, and alert Zach about the trends and events affecting the stocks that matter to him and his clients. With uptrends, he can easily see which stocks are trending in online chatter. He can click into those stocks and get an AI summary of recent market moving events, which he can then directly send to his clients, things like John Dears up 5% today because they got an analyst upgrade from, you know, UBS. Most importantly, he can set instant, highly customizable alerts to be notified about the next big event. Just choose the stocks he cares about, pick the types of alerts he wants to receive from price changes to insider trading, set the frequency

Starting point is 00:39:58 wants to be notified, and we'll send him an AI summary via email about the chatter when it matters. So ultimately, what used to take him hours now takes him minutes. Uptrans makes it 10 times easier to stay ahead of market moving events and find the answers he needs right away without any

Starting point is 00:40:13 doom scrolling or fomo required. Uptrans operates as a premium monthly subscription. Anyone can get started for free, and then we have premium plans for more customizable, higher volume alerts. We have a $15 essentials package for DIY portfolio managers and a $50 pro package catered towards investment advisors like SAC. Now, there are 300,000 investment advisors in the U.S. today, along with millions of DIY portfolio managers and retail investors.

Starting point is 00:40:43 So for us to get from here to 10 million in ARR, we need to get to something like 16,000 advisors on our pro plan. And to get to 100 million in ARR, we need to get to 166,000 advisors. Last but not least, our team consists of myself as CEO and my co-founder, Sam, as CTO. Let me say that again. Last but not least, our team consists of myself as CEO and my co-founder, Sam as CTO. Sam and I have 10 years experience as stock market investors. Together, we've written peer-reviewed research on the relationship between new sentiment and stock market outcomes.

Starting point is 00:41:14 I've previously been a financial analyst and Sam was employee number one of a 10-million ARR startup. We're rounded out by our PhD machine learning lead, Joe, and our front-end developer, Hamsa. So that is Up Trends AI. We're on a mission to save investment advisors from the news to help them build better relationships with more clients by staying ahead of the chatter when it matters. Thank you. Okay. So there are great job, by the way. Overall, the pitch is tight in that it explains to me what you do, who your customer is and what the product is. So when you do these pitches, especially in a condensed format, two or three minutes, you really have a very small

Starting point is 00:41:56 number of boxes you need to check. This is not a 30-minute presentation. This is a three-minute or less presentation, which is, you know, to be honest, all an investor needs to start a conversation. Okay. The customer of this product is a financial advisor. And there are registered investment advisors, there are wealth managers, there are financial advisors. There's a lot of different categories here. but it's for somebody who manages another person's portfolio. And it's a B2B to C type product. There's B2B, there's B2C. You're enabling a business to talk to a customer.

Starting point is 00:42:33 In the same way Shopify is. You would say Shopify allows somebody selling stuff on the internet to then reach customers. Fantastic. These businesses tend to be great because you're enabling of an existing business to do more business, to do business more efficiently, or to save money, one of those things.

Starting point is 00:42:54 And so here, you've identified a problem. Do I think you've identified a big problem? I'm not sure yet. Your advisors will tell you, but we know that these advisors get paid a lot of money, right? What does the average wealth manager make in the United States? What is the median wealth manager make? What is their compensation per year?

Starting point is 00:43:15 I know each client for them is an average of $10,000 in the door. Okay, every year. Yeah. Okay. So if they have, but if they have 100 clients, that's a million dollars a year. And you know what? I see these guys and gals who are, and they come at me all the time, Silicon Valley guy, I'm a whale for them.

Starting point is 00:43:37 But, you know, they're going after also my mom and my dad. You know, there's somebody who helps them. And so, you know, people have a retirement account. They got some, you know, 401K and they got some equities, you know, but save the money, you know, and these guys make one percent of it. One percent of a million dollars, ten thousand dollars. Got it. Okay. So, and there's lots of millionaires.

Starting point is 00:43:58 That's growing because equities are growing. And America has a large number of millionaires. UAE has the most imported millionaires right now in terms of where millionaires are flowing. So I do think, this sounds crazy. If you want to raise money, one of the last. the easiest places for you to raise money is to move to Abu Dhabi or Dubai and put your company there because that's like kind of the new Hong Kong or New York and you can get a golden visa and you can get them to invest 250K, 500K out of the gate, boom. They do that for like almost

Starting point is 00:44:30 any American or European or somebody from Singapore, Australian, that Indian that comes and puts their company there. So I'm just going to put that as a little caveat there because money is moving to Abu Dhabi specifically and then also Dubai to amazing cities. Let's put that on the side here. I think the product looks okay. I think it needs a bit of a design refresh.

Starting point is 00:44:54 It's a little bit too techy and not finance. I want to just talk to you about design for a second. I actually have a question around that. Yeah. Okay, tell me your question. So speaking of design, I'm thinking a lot about our team right now.

Starting point is 00:45:06 We're a team of four getting ready to fundraise. Yeah. And you've talked a lot in the past about the importance of your founding team. But thinking about what's next, my question is like those two to three next specific job functions or hires, where should we be focusing? Well, let's talk about the four you got.

Starting point is 00:45:24 I'm hoping two of them are writing code. Four of us are writing code. God, I'm in love it your company already. You got four people writing code. How many of them are founders? How many of them are employees with stock? Two founders, two employees with stock. Perfect.

Starting point is 00:45:37 That's great. So you got some redundancy there. you can basically there's going to be two more positions you have to add at some point. One is going to be somebody to do sales. And that person right now should be one of the founders. Why should you do founder-led sales? Because you need to know. You need to have customer zero, customer one, two, three, and you've got to be able to listen to them.

Starting point is 00:46:03 So let's pause for a second here. Tell me about how many paying customers you have and How much you charge ballpark and how do you charge? Yeah, so right now we have a few hundred paying customers, two paid plans, a $15 a month and a $50 a month. I will say originally we were more focused on the B2B retail investor. And after feedback, we've learned a lot about focusing on the pros. You made the Cardinal sin. Yes, we did.

Starting point is 00:46:27 Of all startups. But we've learned. You tried to run two different businesses concurrently a B2C and a B2B, but you figured out that B2B is the next Cardinal sin, which is you are charging far. too little. We just established that one customer equals $10,000. Do you believe you will keep, you will get your clients, these wealth advisors, do you believe you can get them one extra customer a year? I think so. Yeah. Okay. Do you think you could save them from churning a customer every year? Easy, yeah. Okay. What is the value of getting a new customer and not losing a customer to them? Yeah. I mean, that's my pitch right there. What is the value, though? I'm asking you a specific question,

Starting point is 00:47:07 a dollar amount. Well, if it's $10,000, $1% commission. Yep. Whatever that is. $10,000. And they would have lost one that cost them $10,000, and they would have gained one that's $10. So you've created $20,000 in value, which means the LTV. Do you know what that stands for?

Starting point is 00:47:25 LTV. Yeah, left time value. Perfect. Okay, I'm just benchmarking where you're at in your startup journey. Your LTV is people will stick with this product for seven years. I'm going to guess. Maybe five on average. Let's pick five and be conservative.

Starting point is 00:47:37 That means each customer you acquire. is worth $100,000 to you. That means your CAQ. What does CAQ stand for? Customer acquisition costs. Perfect. Your CAQ could be $1,000, and you would make it back very quickly. So the value you're providing, if we blowball it, you gain one, you don't lose one, you know, $20,000 a year.

Starting point is 00:48:02 You're providing in five years $100,000 of value. That means you really should be charging 10% of that number, which is $10,000. $1,000, which is $2,000 a year. You're charging $50 a month. $50 a month, you know, is but $600 a year. So you probably for this product, should be charging $500 a month, $400 a month. Because you can really justify it. That first customer you save, you should get 100% of it in my mind.

Starting point is 00:48:30 Okay. You should get 100% of it. So that would be $1,000 or $800 a month. So I would just get rid of all this pricing and be taken seriously by your customers. a wealth manager is spending on lunch with their client $600 or dinner. They're taking them to a Knicks game or a Mavs game, and they're sitting in the first four rows for $10,000. That's how they think.

Starting point is 00:48:53 And you're coming to them asking for $600. It's like they pay more for, you know, that's what they're paying for their Gmail account. Come on. Sure. Yeah. It's a signal. It's a signal. On this, right?

Starting point is 00:49:06 Signaling, yes. So signaling is way off. Now, what this will also do for you is it's going to have you capture the high end first and then go down stream. You want to capture the high end because the high end is going to have the best advice for you. So not only by raising your price, do you increase perceived value, you have more money to hire people, and you have the ability to get the best chef's kiss best advice. So I want you laser focus, not on the number of customers. I want you to get, I would rather you have the 10, 10 of these wealth advisors who make

Starting point is 00:49:41 $5 million a year or $2 million a year than for you to have 200 of them that make $400,000 a year. You want to go for the really high end here. And they're going to give you great advice. So I think you've got to redesign the product at the U.S. a little better. And I think there's some virality here that you haven't thought of. And this is what a jam session is about. You identified like there's a problem. Oh, they send the customer, customer says, oh my God, I own Tesla stock, Uber stock. Oh, my God. Uber's up like four bucks right now.

Starting point is 00:50:10 I just noticed before I got on air because the Robo Taxi is being delayed in some way. Who knows what's going on? And now you've got this like existential thing. I mean, this is like panic-inducing for Tesla shareholders, Lyft shareholders, Uber shareholders. If I had exposure to that, I do. I would be like, oh, my God, I'm not. But anyway, putting it all aside, be really interesting if when you share. shared the dashboard of the stocks with a customer.

Starting point is 00:50:36 If the wealth manager got a ping, Jason just opened the website. Jason, just like a docu sign. So look at the docu sign, hey, the customer opened the contract. The customer went to page two, where you send your deck to somebody and like whatever the slide deck thing that watches. Oh, they stayed on, they went back to slide five. You know, they spent two minutes on slide five and they just zip zip fast six, and eight, why? Oh, six, seven, and eight are irrelevant to them, but they really cared about

Starting point is 00:51:05 the team slide, but they didn't care about the go-to market or vice versa. So you got so much you could do here in intelligence, and people will pay for that, big time. If I know that my customers are typing in the ticker symbol Uber, and then if they could write a question on that portal where they said, you know, let's say, you know, I'm the, I'm the, I'm the, I'm the I'm the wealthy individual. You're the wealth advisor, Ramsey. And I'm on the website and I just say like, what is this about?

Starting point is 00:51:35 Why is Uber up and Tesla down if Tesla is going to kill Uber according to this story? And then you wrote back, okay, here's the Goldman Sachs report. That's a business insider story. Business insider is sensational. They want to get you to click. Goldman Sachs has an analyst who's been covering Uber. This is the analyst's name. And I direct you to that story.

Starting point is 00:51:56 Now the interface has the intent. higher history of us going back and forth. And then additionally, this all has to be mobile at some point. So getting a mobile and a designer is critical. And then I think the next piece would be to have what's called an SDR or business development rep would be very good for you to have somebody trying to figure out who wealth managers are and a viral way to get them on the phone with you. Those would be my next two or three hires if you could get the money in here.

Starting point is 00:52:24 You're doing a great job. You got a great idea. and I understand you told me you were talking to some of your customers in a group chat somewhere like a Discord or a signal or a WhatsApp? Yeah, just iMessage. I messes. You got like a you have one to one

Starting point is 00:52:38 discussions or do you have like a product council yet? We've got a discord for like large group and then I have like a small group of three or four advisors that I text weekly. Okay, awesome. That's really what you want to do. Ramsey, I wish you massive success with this idea.

Starting point is 00:52:54 It's supposed to be just 10 minutes, but your business is so great. You're at 16 minutes with me. You jammed with JCal. Rate your jam with JCal. How would you rate this in terms of helpful? Huh? Give me an honest number between one and 10. That was great. I would say 10. Yeah, for sure. Okay, there we go. I don't want to bias it in anyway. But you, you know what? You should meet our team. So you might be a good candidate to come to our accelerator because I think there's something here. And when I jam with somebody, I got to tell you, you're good to jam with. Because you're quick, you give really good answers. You don't filibuster. And that's a what people love when they're jamming. So that's a really good note for everybody listening to

Starting point is 00:53:29 this week in startups. When you're interacting with an investor who's a know-it-all investor, who's seen it all, who's invested in hundreds of companies, you've got to be able to go back and forth quickly, right? And you've got to be able to have that real intellectual discussion, and you did good with that, right? It's like kind of playing pickleball or ping pong or tennis. You want to have a good volley. In just 15 minutes together, we had a great volley. I like you. I like the way you answer questions. I like the way you think. You kind of did your thing. You made yourself likable. I understand your business.

Starting point is 00:53:58 You're thoughtful. You seem like you're, you got a chip on your shoulder and you want to be successful. I think lean into that a little bit, like a little bit of the drive. Don't be too mellow. And congratulations, Ramsey. And we'll see you all next time on Jamitjow.

Starting point is 00:54:14 Thank you to our sponsor.com.

This Week in Startups - The Power of WEKA with Antonio Gracias and Liran Zvibel + Jam with JCal: Uptrends AI | E1986

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