FULL TRANSCRIPT

‍

‍

Speaker 2 (00:00)

Okay, so let's get started. Super excited for today's conversation. My name is Slava Rubin. I'm one of the founders here at Vincent. We're your platform to be able to educate you and learn all you need to know about the alternative investments markets. We have Jan-Erik with us from Sacra You wanna say hi?

Speaker 1 (00:19)

Hey Slava and yeah, thanks for having me. Great to be here as always.

Speaker 2 (00:23)

Absolutely. I don't even know how many times we've done this now, but you know, we should take the show, you know, in real life one day. We're not quite the all in podcast, but you know, one day we'll get there. So today's conversation is not focused on one pre IPO company, but rather some of you have asked, listen, I'd like to just get a few at a time to be able to digest it a little bit quicker.

Speaker 1 (00:30)

I know.

Speaker 2 (00:47)

So today we're gonna be covering five awesome pre-IPO companies you need to be knowing about. And today's conversation is brought to you by our partner, which is the Fundrise Innovation Fund. So let's just dive right in.

So the first thing to know about, and we actually have news hot off the presses today, is everybody is trying to figure out how do I get into these pre-IPO companies. Getting into the public companies is obviously quite easy. You just get your Robinhood account or you get whatever brokerage account that you use, and we've now gotten pretty used to trying to buy some Nvidia stock or some IBM stock or Apple or whatever it is.

But when you hear about SpaceX or when you hear about Anthropic or you hear about Stripe, people keep on wondering like, what is my way of getting into these deals? And that's obviously one of the things that we're trying to educate everybody here at Vincent. So we just wanted to break it down just a little bit. Many of you are already familiar with this, but we just wanna make sure that we're all on the same page. And just as a reminder, we do have the Q &A at the bottom. So if anybody has any questions they'd like to ask,

feel free to list them at any time, including right now. I will do my best to review any of the incoming questions and try to incorporate it as appropriate into the conversation. So there's really four major ways that one of you that are listening to this right now can really enter into this world. First is there's the Reg D offerings, where you could try to participate directly into around on the cap table.

You could try to get 25,000 minimums. You could try to get even lower. But often, this is gonna be tricky because you need to have access to the direct deal. Not through an SPB, and we'll talk about that in just a second, but rather your money going directly on the company's cap table. Some of the companies that are offering things like this are Equity Zen, Republic, and others. Now what's happening a bit more frequently is

some manager is pooling money on behalf of many investors into an SPV. An SPV is a special purpose vehicle that is being created for the single purpose to get onto the cap table of one of these companies. The reason these SPVs are needed is often these larger companies, let's just take for example Stripe, is they don't want to have

hundreds of additional one-off investors that they need to then deal with for a compliance perspective and then from an investor relations perspective. So sometimes 10, 50, 100, whatever it is, investors will all be rolled up into one SPV, one pooled vehicle, that will show up as a single line item on the cap table. And then as it relates to the company, Stripe, they will only interact as a compliance perspective and for a investor relations perspective.

with the manager. Now that obviously has its own issues compared to being on the cap table directly. And you should reference some of our other content where we actually break down what is it like to be able to get various deals or various access into these companies. We did an entire one hour session covering all this content and all the terms you need to know. One of the more innovative new approaches is these 40 act funds. So very...

compliant, innovative, progressive concept, which is these are SEC regulated vehicles with liquidity provisions. So this is a bit of a hybrid. You have really almost like a public vehicle that is then able to take on private investments inside of it. So one of the more well-known one is the ARK Venture Fund. So we actually had Cathy Wood recently talking here with us at Vincent.

and she was covering some of the beauty of the ARK Venture Fund. And there's also another example, which is Fundrise, which is our partner today and could be an opportunity for those who are listening on this call. And finally, there are these marketplaces, Equities and Ogman, Zambado and others. And this has been around now for a while. This is almost like a mini NASDAQ or like a mini...

Dow or New York Stock Exchange. Now, they're not regulated the same way, so you definitely need to make sure you're looking at the shares appropriately, understanding where you're buying, where you're not buying. But they're trying to create a similar liquidity for these private shares. One of the most well-known ones, EquityZen, which I mentioned earlier, actually was just bought today, acquired today. Who's that by, Jan-Erik?

Speaker 1 (05:34)

Morgan Stanley.

Speaker 2 (05:36)

So Morgan Stanley actually acquired Equity Zen literally today. So I don't know if they knew we were gonna have this discussion. So Morgan Stanley timed the acquisition for the webinar or if it's the reverse, but you can decide on your own. The key to this is actually, it just shows that the legacy players, Morgan Stanley is getting much, much, much more into

like the dynamics about what's happening in the private markets. Another example of this happened just a few weeks ago, a little bit less press, well, maybe not less press, maybe a little bit less known of a company, which is Industry Ventures, which is a VC fund from Silicon Valley that also has a fund of funds business that invests into other funds, as well as a direct business where they try to get direct into really interesting private companies.

They already had billions of dollars of AUM and Goldman Sachs actually bought them for nearly a billion dollars. And there's a lot of coverage recently about why Goldman Sachs was doing that, trying to get more into the private market opportunities. And now you have Morgan Stanley buying EquityZen hot off the presses earlier today. I mean, this is very timely what we're talking about, which is everybody from a business perspective or an individual perspective is trying to figure out.

How do I get more exposure to these exciting private companies as opposed to only the public companies? And there's a couple main reasons about this. One, 30 years ago, you had over 8,000 public companies. Now you have just five and a half thousand public companies. So there's less public companies, which is crazy 30 years later. And two, the public companies now are becoming much bigger before they go public, meaning the private companies. It used to be that you IPO.

20 years ago at 500 million on average and now it's 5 billion on average and that's just the average. Nevertheless, these huge companies, the OpenAI, Anthropic and others that are well over 5 billion, if not hundreds of billions of dollars. So that's why many of you are listening to this conversation. Jan-Erik, anything else you want to add at the moment around these different pre-IPO pathways?

Speaker 1 (07:49)

No, think you said pretty much everything there is to say. Great overview.

Speaker 2 (07:54)

Perfect. I love the compliments. That's why we keep you around. All right. So let's just dive right in to our companies. So today we have these five exciting companies. You probably have heard of most. It'll be varying degrees as to how much you've heard us talk about any of these. So I'm sorry if there's any repetition for those in the audience who might have heard our one off sessions where we do a big deep dive into any one of these companies. But we thought we would do more of a cursory review of these five companies.

We've had feedback that this is interesting, so we'll probably be doing more of this into the future as well. With that said, I think we should just dive right in to, so just to know what the companies are, Databricks, Groq, Anthropic, Anduril, and Canva. So let's dive in to the first company.

Let's go with Databricks to begin and take it away, Jan-Erik.

Speaker 1 (08:48)

Awesome. Yeah, Databricks, you know, it's kind of one of the biggest private companies now, $100 billion valuation. One of these companies that people expect to go public. Sort of the biggest competitor or, you know, comp in the public markets is Snowflake, which obviously had a giant IPO back in 2020 and has done well since then. But it's also one of the least understood. So I think it's a good one to start off with.

To go back to kind of beginning, think it's helpful. Mid 2000s, you had companies like Netflix blowing up, you had company like Amazon doing really well, all building these kind of personalization algorithms so that they could recommend you the exact widget to buy or movie to watch next. And this whole concept of big data was really nascent. Facebook, Yahoo, giant companies like this.

The sort of key insight that Databricks had was around doing these kinds of large scale machine learning experiments required, having a giant infrastructure data team inside your company already, because the sort of software frameworks that existed for doing this required a lot of manpower. And what happened is you had a team of researchers, UC Berkeley, kind of classic Silicon Valley,

know, science experiment that became a company, which was basically a, you know, open source engine that would allow teams to do this kind of machine learning work much, much faster. Well, that was the early iteration of the company. By 2019, they start to ride this next wave. So big data was the first wave. The second wave was kind of multi-cloud and really the huge migration to AWS, GCP, Azure.

And what Databricks did is that they launched Delta Lake, which has become a really core product. It was sort of a rethinking of the traditional data warehouse, which sort of brought this core use case of doing much faster analysis over data, being able to build ML algorithms to do stuff like personalization recommendations. And they really kind of...

came into their own as a competitor to things like Snowflake and Amazon Redshift in 2019, 2020. That was sort of the next growth phase, which took them up to a billion in revenue. Plus, and at that point there was kind of an inflection where Databricks could have just become kind of a fighting Snowflake in the trenches to comparable-ish companies. But fortunately for Databricks, they have had a third wave emerge.

which is generative AI. So November of 2022, chat GPT comes out. Shortly afterwards, know, every enterprise in the world realizes that AI is a huge thing that they need to do and get good at. And Databricks, you know, has been building for a decade. It's a vertically integrated infrastructure around, you know, big data and also training ML models over that data. And

they were extremely well positioned to capitalize on this, which they did by acquiring Mosaic ML. They started launching their own proprietary models, large language models in 2023. They were one of the first or maybe the first company outside of the foundation model labs that was doing this. And they have sort of fully pursued this. grew to, ⁓ their AI products grew to 1 billion ARR basically within one year and they just crossed 4 billion.

in August. So that's sort of the overview. I'd say the one key metric or stat that's interesting about Databricks right now is that they are almost as big as Snowflake. Snowflake is still around, I think, 4.5 or so. So still bigger, but Databricks is growing almost twice as fast. So 50 % year over year versus Snowflake at roughly 27%, 30 % year over year.

you know, on route to pass it then 2026 at some point.

Speaker 2 (12:43)

Right, so it's interesting because this company is now worth $100 billion in terms of their last valuation. And that was a big jump versus I believe it was like a 2 billion or something valuation just a little while before that. Is that right?

Speaker 1 (13:01)

I think they were at, you know, I think a few years ago it was around 30 something around there. But yeah, it was a huge jump that happened pretty recently.

Speaker 2 (13:13)

Right. So the hundred billion by itself seems like it's quite expensive because it's just a big number. But then when you start comparing it to Snowflake, it becomes really interesting, right? Because Snowflake today, if you look at the market cap is 91 billion, right? I think that's at the moment 91 billion, which that's the public company, right? The public company Snowflake, which is doing similar work is around 91 billion.

And it seems as if their multiple is also lower, meaning the public company, because they have a little bit more revenue, you mentioned. How much is that the Snowflake has?

Speaker 1 (13:48)

around forward-looking revenues roughly four and a half. So the multiple for Snowflake, I believe is about 18, 20X, somewhere in there.

Speaker 2 (13:56)

Right, so it's interesting. So the question to ask is, you know, is Databricks, if they're apples to apples in terms of competition, is Databricks too expensive because it's just private, so it's just getting a private premium. And if it was public today, it would all of sudden be trading at an 18x multiple like Snowflake, or is the counter that they're just growing significantly faster?

And really, they might be underpriced. And if they were public, potentially, Databricks actually would be quite a bit more expensive, where people would be chasing the growth. And maybe some of the snowflake investment dollars would be moving to Databricks. What are your thoughts on that?

Speaker 1 (14:36)

Yeah, I agree. The surface read for me is definitely that it's underpriced given that it's growing twice as fast and has really has caught up to and is, I would say on track to eclipse Snowflake at the current growth rate in the first half of next year. So that alone, I think would suggest a 25X versus Snowflake at 20X is pretty justified and potentially, as you say, underpriced, especially with

know, gross margins of 85 % versus Snowflake at 70. I think it's a compelling story. I think the flip side and maybe why, you know, why you might weight that down a little bit is because, you know, one, there's this concern in the public markets right now about is there an AI bubble, right? And folks have a lot of strong opinions on either side of that. But I think that that could be one reason why.

Databricks hasn't gone public and they've talked around the idea and it seems like they probably will, but I think that that's one factor that could lead to some hesitancy with investors thinking that if Databricks has doubled ARR since the launch of Jad GPT, mainly because of folks enterprises paying them for AI, then any disruption to that narrative.

that AI is taking over everything could be very bad for their growth rate. So I think that's one big one. I think another one is, Snowflake has had a very salesy culture for a long time and they have a pretty longer track record on the public markets. Snowflake has had a professional CEO in charge for a long time. Databricks is a very different kind of company.

One of the co-founders, Ali Godzi, is still around as the CEO. It's a very academic culture, engineering first culture. Obviously, they've grown to 4 billion ARR, but they're transitioning into kind of more, trying to be more sales-driven, you know, enterprise kind of company that you would see on the public markets that would be a little more, you know, relatable to Wall Street. So I think there's another question there. And there's some suggestions that, you know, the sales organization is struggling with this.

you know, transition, you know, DNA kind of evolving, which is natural, but just another thing to look out for as they, you know, potentially go public.

Speaker 2 (16:55)

As a reminder, if anybody has any questions about anything or specifically Databricks, just jump into the Q &A and ask your question. So I believe Databricks raised like a billion from Thrive Capital. So they should be, recently, they should be fairly well capitalized. And you mentioned the concern around potentially the AI bubble for the public markets. You know, this is not financial advice. You don't know any insider information, but

Would you predict that Databricks is going public by the end of 2026?

Speaker 1 (17:27)

I think if things continue to go well in Google, NVIDIA, et cetera, universal companies, and that narrative is still working, then I would say yes, why not? But I think, and then add Microsoft in there and OpenAI. I think if the kind of

If there are indications that there's an AI bubble or the AI bubble starts to deflate or pop, then I think they could easily, you know, with the billion dollars they just raised and with being cashflow positive, they could stay private. So they do have that option, I think, you know, as, as, as, yeah, which is good for them. But I think otherwise I could see them definitely going public by the end of next year.

Speaker 2 (18:16)

So let's say you had access to Databricks right now and you could buy it through the private markets. Obviously, there's going to be some fees associated with that one way or another. We don't know exactly what those terms would be. Or you could buy in the public markets, you could buy some Snowflake. It seems as if maybe superficially Databricks is a bit more exciting at the moment. But Snowflake

definitely is the easier, more transparent user experience to get great pricing on the investment and also to do it quite easily. Do you think it's necessary to buy Databricks to get exposure to this? Can you just get it through Snowflake or is that just too simplified? Really, Databricks is a very unique N of one company.

And it's worth going through the effort to try to track it down and get it into, you know, your basket or into your individual stock, you know, exposure or into one of these funds.

Speaker 1 (19:13)

Yeah, I don't see Snowflake as quite as indexed on AI. And obviously they're repositioning hard around it. But I think to get exposure to enterprise adoption of AI, I think I would rather look at things like, well, I would rather look at Databricks. And in the public markets, probably things like Palantir, I think is kind of riding the same trend as well as your Nvidia and your Google.

Probably more so than a snowflake.

Speaker 2 (19:46)

All right, so as you just said that sentence, five people just went to go buy some Databricks. Databricks is closer to Palantir. I imagine that that's the way to create a meme.

And then let's move on to the next company.

Speaker 1 (19:56)

Groq Yeah, so Groq will be an interesting one because out of all the companies on our list today, they are by far the smallest in terms of revenue, by far the sort of least developed and well-known company, but doing something very interesting that a lot of people are excited about. So basically Groq is in this kind of category of AI chip companies.

You might have heard of companies like Sanbonova Systems as well. There's a new one called Unconventional, which is founded by the former co-founder of Mosaic ML, which Databricks acquired. There's also Cerebris, which was supposed to go public earlier this year and delayed that. So there's a lot of these companies all kind building these alternative chip systems to what has been offered by NVIDIA and

Groq's kind of special sauce or special use case is all about speed and low latency. So that's what they've been building is these special chips that are not GPUs, which basically enable large language models and other types of AI models to be inferenced at very, very high speed. So if you think about things like a voice agent where you can speak to it,

and it will reply to you with zero latency rather than like an open AI voice mode. You talk to it, there's always like a second or two that it takes to think of a response. Groq chip voice AI is basically instantaneous like you're having a conversation with a human. Or you can look at things like if you had a humanoid robot working at a factory.

you know, would it have the ability to kind of take actions almost in real time based on what it's seeing or does it have to think, you know, does have to sort of have that pause? Other kinds of real time apps, you can imagine this real time speed is very valuable. So that is the essence of what they're building. They had $90 million of revenue in 2024. They are projecting 500 million in 2025. A big amount of that right now is based on

Groq's work with various countries in the Middle East that are building out sovereign AI systems. So one is, they're building out the largest data center for AI in the world, apparently, in Saudi Arabia. And they're helping ⁓ Saudi Arabia also with their kind of sovereign LLM. But they're also doing work with various national laboratories, things like that.

⁓ which is sort of the main source of revenue now versus what they imagine is eventually chip sales will be the main thing and cloud inference via Groq's website, Groq's API. To cap it off, think the main kind of downside or the main risk factor right now is that while the chips are very fast, very, very low latency, they do not handle

currently, large amounts of concurrent users. The cloud service doesn't handle large amounts of concurrent users. And then right now they don't handle particularly large models super well. So if you're trying to train the best frontier models right now, and you're trying to serve them to a large number, millions of people, you are still using NVIDIA GPUs or Google's chips. And those are basically the two that are out there. And so Groq,

there's still some sort of technical hurdles to be unlocked for them to work at scale. But the long-term thesis is basically that there are these use cases where real time is going to be essential. And it will never be acceptable to have a sort of delay or have any kind of, whether it's 100 milliseconds or whatever, there's never going to be sort of room for that.

And that will be a very large and valuable market for Groq to own.

Speaker 2 (23:39)

If their product is so strong, why is it that OpenAI, Anthropic, Oracle, AWS, Microsoft, others are not buying from Groq yet? I it's supposed to be faster and cheaper, right? So those are two great things. Is it the thing you were just saying? Can you expand on that?

Speaker 1 (23:58)

Yeah, essentially. like open AI, for example, is also working on their own chips with Broadcom and that would replace their potentially, you know, augment or replace their use of Nvidia chips. But the whole point of what they're building is to be able to deliver, you know, chat GPT, voice mode, video generation, you know, ⁓ to at this point, you know, hundreds of millions of people around the world, you know, eventually billions and

For that, the GPU is sort of the best in class, state of the art, if you wanna be able to deliver to that many people. And speed, from what they've found, is not really the core limiting factor. The other factor is cost. So like, if you are talking about building a chatbot that would deliver,

text back to a user, the hardware cost of getting Groq chips, and then the sort of monthly operating costs versus getting Nvidia GPUs, it's like 3 to 4, or sorry, more like 10x the cost. It's like 3 to 4x as fast, 10x as expensive.

Speaker 2 (25:08)

Sorry, it's more expensive. Groq is more expensive. Yes.

Speaker 1 (25:12)

Yeah, sorry. It's more expensive to deliver. It's just much, much, much faster. Got it.

Speaker 2 (25:17)

And what do you think is the unlock for them to actually get these massive clients like the OpenAI, Anthropics, Oracle, AWS, Microsoft, et cetera, which then sends this company to potentially 100x devaluation?

Speaker 1 (25:34)

Yeah, I think it basically hinges on them continuing to figure out how to scale the model and get it such that, like, I'm sure if OpenAI could call upon Groq to deliver enough chips that they could give people super low latency voice mode. That stands to reason that could be a very compelling experience. The problem is right now,

that it's just not possible to actually bring that into being on the architecture side. Groq still has that to figure out. I think it's possible. think probably these deals like these Middle Eastern sovereign AI deals are allowing them to not only generate revenue, but start scaling up on the infrastructure side, building up these data centers. I think that that is a potential avenue like road to

being able to get these bigger customers. And we have seen the glimmers of this kind of customer acquisition happening. I know that Perplexity uses Cerebris chips in some capacity. It's not quite clear exactly what percentage of Perplexity queries will go through one of these Cerebris chips, but I can imagine Groq starting to get similar partnerships like that going so that they can sort of get experience.

experience doing that. But yeah, for now, it's more on the on the sort of government side more on the research, you know, science side. And we'll have to see next year, see what happens on the on more commercial side.

Speaker 2 (27:03)

mean, Groq literally about a month ago just raised new money, right, at a $6.9 billion valuation, a pretty significant step up. And they had some pretty killer investors. What do you think about Groq at $6.9 billion?

Speaker 1 (27:19)

Yeah, I mean, it's expensive, right? It's like, uh, compared to 90 million of revenue in 2024, it's about a 76, 76 X multiple. Um, you know, Nvidia is at like 20 or 25, um, thereabouts. So yeah, price at a premium. think the main, you know, the main kind of incentive is that, is that if it works, um, then you have, uh, one of the only companies that

can be compared to NVIDIA in the future, assuming a world where AI continues to be very important and increasingly so every year, Groq becomes kind of like the real time, low latency chip provider of choice and they could own 10 % of the AI chip market, conservatively assuming that that's the market for voice AI and robotics and that kind of thing.

NVIDIA today is a $5 trillion company. So if Groq can be anything comparable to that, then paying $7 billion today is nothing. So I think, yeah, definitely, potentially could be very, very big. And if not, it's a total loss probably, but that's probably what they're thinking.

Speaker 2 (28:35)

And then there's discussion that they're moving more into data center to try to get more of the Crusoe core weave kind of exposure. Any thoughts on that?

Speaker 1 (28:46)

Yeah, Groq data centers, like I said, yeah, they're building one in Saudi Arabia. And I think, you know, sort of riding this, bunch of these countries in the middle East want to become kind of, you know, want to bring in a lot of data centers and leverage some of these vast kind of energy resources that they have and to stay sort of, you know, stay modern relevant with the changing economy. I've seen from Groq that they want to establish

you know, 12 plus new data centers next year. And a lot of that focus is going to be in the Middle East and in Asia. So yeah, definitely interesting. And we'll see sort of, I think, next year a lot more on that front.

Speaker 2 (29:26)

So just going back to Databricks for a second, if you were a buy, sell, hold, green, yellow, red light, it sounds like you were a buy, green light on Databricks, is that right?

Speaker 1 (29:37)

Yeah, I mean, I think I'm a little more maybe bullish on AI than the median person. And I think that they're really well positioned there. So, yeah.

Speaker 2 (29:48)

And then similarly here, know, Cerebris, which is kind of similar is having some challenges on trying to go IPO. At the same time, you're seeing a ton of consolidation going into Nvidia with now you're at five trillion. So as much as everybody's saying, you need competition for Nvidia. So it's just obvious that the money will go elsewhere. I mean, it's not like Nvidia is really slowing down. So

With all that in terms of the positives and the negatives here, are you green, yellow, red light on Groq? Bye, so hold.

Speaker 1 (30:20)

Yeah, yeah, it's interesting. You know, I like Rockets. It's not like I would put my entire, you know, all my chips in there. But I think as far as, you know, AI chip challengers, they're one of the more interesting ones to me. You know, because yeah, like you said, Cerebras had some issues. I know San Benova is looking to achieve a sale because of some issues that they're having with their latest raise.

And then beyond that, there's not really major players. So I kind of see Groq as right now being the best position player in the space. So I'm cautiously optimistic, I would say. Green slash yellow.

Speaker 2 (30:53)

All right, let's move on to the next company. Anthropic, some people might have heard of this one.

Speaker 1 (30:59)

People might have heard. People might have heard. this is so enthropic. If you don't know, it was founded by two kind of defectors from OpenAI very early on who wanted to build their own AI lab. And they've done so. They've built basically the second biggest private AI lab at 5 billion ARR to OpenAI's 13 billion. They're really their explosion of, you know,

Success came in 2024 with the release of their Claude Sonnet 3.5 model, which has since been updated to 3.7 and then now currently 4.5, which really became the backbone of all of these vibe coding and AI development tools being first kind of cursor, which is now 500 million ARR and then a whole range of tools more for non-technical folks.

Bolt.new, lovable, replica agent, and a whole slew of those kinds of tools, all of which relied upon Anthropic's models basically exclusively because they were simply the sort of state of the art for coding, which has remained true basically, yeah, since the summer of 2023. And this is why Anthropic has kind of, you know, has kind of become known as the best, you

AI models for coding, they are heavily B2B centric in terms of revenue. So roughly 80 % of revenue comes from enterprise, API usage versus open AI, which is much more tilted towards consumer and these subscriptions to try GPT, which really drive revenue for them. So yeah, you can think of Anthropic as kind of like this B2B and more coding centric version of open AI in the most simplistic way.

And yeah, they're also expanding into finance. So big launch the other couple of days ago of, ⁓ know, Claude for finance where they are working very closely with banks, which are eager to replace their junior employees with AI and Anthropic is eager to help them. So we're seeing that. We're also seeing the other big trend I'll mention is, you know, there was a very kind of testy or, you know, ⁓

interesting relationship between Anthropic and tools like Cursor. As you know, Cursor was kind of offering developers a way to get Anthropic models to be used on their code base. And now Anthropic is directly competing with things like Cursor because they launched Claude Code, which is their own coding tool that developers use. And that, you know, since launch that went...

you know, in something like four or five months to 500 million ARR. So it has really grown into a significant business on its own and kind of goes with what we're seeing across the industry, which is a lot of these labs like Anthropic are launching more and more products for end users and then owning both the sort of model layer and the app layer. So yeah, quick overview.

Speaker 2 (34:01)

Yeah, so obviously they're competing with OpenAI and it seems as if their revenues are growing quite substantially as you can see by the chart. The multiple that you have here, the 36.6, is that on the $5 billion ARR? Is that the multiple you're using?

Speaker 1 (34:23)

Yeah, 183 billion valuation and five billion error.

Speaker 2 (34:27)

Got it, got it. And then by the end of year though, they're supposed to hit 9 billion. And do you really believe there's 20 to 26 billion by the end of 26? I mean, that'd be pretty substantial.

Speaker 1 (34:37)

What was it? know. Yeah. Yeah. Twenty twenty six. I think. Well, yeah, I think it's possible. mean, a lot of these companies have projections, but OpenAI also puts out projections and they're all kind of hard to believe. And then you look back at the last three years and that would have been this would have been hard to believe back then. I think they're projecting based on current trend lines. So we will see.

Speaker 2 (35:03)

Right, so I guess just trying to do the math here. If they end up doing 9 billion at the end of this year, and let's count their $180 billion valuation, that's basically like a 20x on the actual realized year. If they really hit 20 billion at the end of 26 using a 20x, that'd be like a $400 billion valuation, right, if that actually were to happen.

I'm just trying to think through some of the numbers here. And is OpenAI, does it have similar multiples?

Speaker 1 (35:35)

Yeah, OpenAI most recently raised at $500 billion, which I think makes them maybe the most valuable private company, but they're at about a 38x multiple on $13 billion ARR.

Speaker 2 (35:48)

Okay, 38x, so way higher than 20x. Got it. And where's your, we do have some questions, one second.

One was a comment, is, Vercell is cool also. Agree. Agree. Nice, nice.

so one person here is predicting that OpenAI in three years is a trillion dollar company and in five years is a five trillion dollar company. So this is this person's prediction. Do you think how far behind is Anthropic is their question?

Speaker 1 (36:23)

don't think they're too far behind, maybe a year or two, I would say. Yeah, but I agree. I agree on them getting to one trillion, maybe in less than three years because they're at 500 billion now.

Speaker 2 (36:33)

So if you were to, so that was on OpenAI. You think OpenAI hits a trillion in three years or less. If you were gonna say Anthropic was gonna hit a trillion, A, are you confident of that? And if yes, when?

Speaker 1 (36:45)

I think Anthropic, I think probably OpenAI will hit that. I mean, I don't know, maybe I'm like, I had a monster this morning, but you know, I think probably in like two years and I think Anthropic in like three, three to three and a half.

Speaker 2 (36:57)

Okay, so let's go to the green, yellow, red, buy, hold. If you think Anthropix hitting a billion in three years and you could try to get into it today at $183 billion plus fees carry whatever, I don't know what price you're gonna get into it. Does that mean you're bullish?

Speaker 1 (37:17)

Yeah, I think this probably we're on the most green because they are, you know, the number two AI lab, but they're also, they've proven durability over the last two years and being the best coding model. And that hasn't changed. And I think, you know, it's evidence of some, you know, durable advantage that they have acquired, given how quickly everything else gets commoditized in sort of, in AI training. So, yeah, definitely enthusiastic myself.

Speaker 2 (37:43)

We have a curiosity question. Does Verso depend on Claude like cursor or bolt?

Speaker 1 (37:49)

not so much because they are not a pure play vibe coding platform. I mean, for them, they have V zero, which I agree with the commenter before, you know, Verso is quite cool. It's, it's better for front end. I think it's great for design. Everything looks really nice. And then you can deploy it straight onto Verso. Verso's hosting business is really the key revenue generator. So I think of their vibe coding platform is more top of funnel, you know, a way to get more people to generate sites that get hosted on Verso. but.

You know, I think that they could just use GPT-5 or whatever. I don't think it's as essential for them.

Speaker 2 (38:25)

Got it. And then just to repeat, you said you were the greenest of the green here. So that was very interesting. Let's go on to, which I would totally agree with for what it's worth. Let's go on to the next slide.

Speaker 1 (38:36)

Awesome. Yeah, also another one I'm excited about. Anduril for folks who don't know, if you go back 10 years, defense tech was considered completely uninvestable by the vast majority of VCs because it takes years to sell anything into the military. It's all hardware based, which was very unattractive in the golden years of SAS. And also,

a lot of the recurring or revenue is not recurring, right? You get a one-time sale. So this was not an interesting market. The sort of flip side of that, that Anduril will recognize was that there was a absolutely enormous, you know, addressable market because the Pentagon's budget, you know, trillion plus every year. And also it's dominated by sort of Lockheed Martin, Boeing, Raytheon companies like that.

that are, you know, have been around for a very long time and do certain things very well, but also have created a large amount of, you know, they create a large amount of waste and then they're kind of, you know, fat and happy in many ways. So Anduril came in and flipped the business model of defense and said, instead of, you know, instead of signing the government up on these long-term contracts where we get paid basically our entire, whatever it costs us to build this fighter jet,

we're going to get paid whatever it costs and plus a little more. Anduril they said that they would do R &D upfront, build products that the military would want to buy, kind of take a Silicon Valley approach to defense and then sell those into the government afterwards to save them money. But also, ideally deliver these more advanced, more modern AI products that can be better. And so they started with notoriously

you these border towers on the southern border of the United States. And that was the initial form. And then have just sort of expanded, you know, entered many of these little mini markets within the DoD and gotten into, you know, drones, counter drones have become sort of even bigger than drones in many ways. then, you know, jets, submarines. So really kind of

they are building the kind of Lockheed Martin or Raytheon for the 21st century is kind of how you might put it. yeah, billion dollars in revenue in 2024. think projecting a doubling of that in 2025, last I saw.

Speaker 2 (40:59)

Sorry, a doubling of that. We're expecting 2 billion of revenue in 2025, is that what saying?

Speaker 1 (41:08)

Yeah, I believe so. I believe so. Yeah.

Speaker 2 (41:12)

So the 30.5x revenue multiple is the 30.5 billion valuation from this summer versus the 2024 revenue, is that right?

Speaker 1 (41:27)

Yeah, yeah, exactly. So they're projecting two billion. So you could look at it as a 15x or maybe split the difference and say 22x multiple on the expected revenue.

Speaker 2 (41:41)

Okay.

I mean, they've been acquiring some really big contracts, right? And this revenue can start scaling pretty quickly, not as quickly as just pure software like Anthropic. But these numbers could get pretty big with what's going on in the geopolitical conflicts around the world. Is that fair?

Speaker 1 (41:59)

Yeah, mean, Ukraine has been a huge inflection point for Anduril. On one hand, just in terms of eagerness of people to join, when they were doing border wall fences, ⁓ Anduril was within this very non-sexy category and shunned in Silicon Valley. And the war in Ukraine really changed that. ⁓ And they've become a very hot destination for engineers and folks ever since then.

So yeah, they've had a huge upswing. think the drone business and the anti-drone business have both exploded since then as well. Not so much supplying to Ukraine, even though you do see Andoril products being used in Ukraine. The bigger thing is kind of the threat of drone warfare has been made very real. And not only the general public, but Pentagon brass,

the executive branch understand that drones are basically the next wave of how warfare will be done. And America needs to have both the offensive capabilities there, but more importantly, probably the defensive capabilities because there's such an asymmetrical, know, anyone can have a cheap drone and attach a grenade to it or whatever. So it's become a huge kind of a huge shift in thinking that has really benefited Anduril.

in a huge way. And then you'll add on, you know, Trump winning the election, he was poised to make one of the Anduril co-founders, you know, one of the top sort of defense administrators in the country. There has been a, you know, a crazy amount of lead over from the Silicon Valley defense establishment and Anduril into the White House. So there's been some sort of tailwinds there as well.

Speaker 2 (43:38)

So in the AI market, you know, the really well known companies are OpenAI and Anthropic, and there are quite a few others that if you're into the private markets that are already becoming very interesting and investable. In this market, there's not really a lot of companies that are known yet. Do you think that there's significant competition for Andrel, like Shield AI or others? And where do you think this heads into in terms of competitive marketplace in the next two years, etc.?

Speaker 1 (44:08)

Yeah, I think there's a lot of interesting companies in the space and Shield AI is one that we've covered. There's Saronic, SailDrone, companies that are doing kind of bits and pieces of the ecosystem. Anduril, I think it's under mentioned how far ahead Anduril is as a result of being so focused on this for so long, since Palmer Lucky sold Oculus and left Facebook really.

⁓ So they're significantly ahead. I would say, you know, if anything, they are positioned as sort of the next big prime contractor, you know, the next big Lockheed Martin, whereas these other companies are more positioned as, you know, kind like the smaller subcontractors. You know, a lot of them, I think, are potential acquisition targets for traditional defense companies that want to modernize a certain line of business. Whereas I think, you know, Anduril is not

is not an acquisition target, is not gonna be an acquisition possibility for any of the traditional primes. They are sort of trying to build the replacement for those. So I don't see those other companies threatening Anduril. I do see potentially they could be also acquired by Anduril. That's kind of the dynamic that I'm seeing right now. I think there's not room for a ton of big players of Anduril size and type. ⁓

in the market.

Speaker 2 (45:26)

10 seconds on Havoc AI. Do you have any thoughts on them?

Speaker 1 (45:31)

I do not. Sorry. Yeah.

Speaker 2 (45:34)

No problem. then there's some commentary that it's harder to get Android in some of these other companies. You have any thoughts on that? I mean, I think Palmore Lucky is kind of publicly known against SPVs. So I think that might be part of the issue. You have any other thoughts on that?

Speaker 1 (45:48)

⁓ I have heard that as well. do think, fortunately, I saw recently, he's saying that they're gonna go public in the next few years. you know, it's not the one I would have maybe expected to go public first out of all these companies we're talking about, but it could be. It could be that they are public, you pretty soon. So that would be exciting.

Speaker 2 (46:06)

Green, yellow, bicep hold.

Speaker 1 (46:09)

Green, mostly because they're, I think, number one in the space. And I think that they're focused on the future of what sort of defense looks like more so than any other company. So yeah, I'm pretty interested in Andriil for sure.

Speaker 2 (46:29)

And just to kind of prioritize, you said Anthropic was the greenest of your green. So is Andral above or below Anthropic?

Speaker 1 (46:40)

slightly below just because anthropics is a little more, I know, I feel like I know it a little better and it's a little more relevant to, you know, things that I do and I'm aware of, but I still admire Anduril So number two in my heart.

Speaker 2 (46:55)

All right, let's go to Canva and let's try to do this one a little bit quicker. Sorry, let's try to get done by the top of our hour if we can just for all of our listeners. Your content's been amazing. We might have to not do five companies. Maybe some people can give me a vote in the questions if they want five companies or less. All right, if anybody has any questions, feel free to do that. Otherwise, on our Canva.

Speaker 1 (47:21)

Yeah, Canva is great. mean, Canva started off as really this like social media image generator, helping marketers put text on images and throw it up on Instagram. And it was a very kind of simple business. And it has just exploded beyond what I think most people would have expected to the point that they're projecting 4 billion ARR by the end of this year, up from 3.3 billion now. And it's really become this complete suite of browser-based design tools for

creating just about anything you can imagine as a marketer. And within the last year, they've really layered on AI in a big way. They acquired a foundation model company that's helping them generate images. They've also built out their own vibe coding platform into Canva. So if you're a marketer, want to build a website, maybe you want to build an interactive ⁓ calculator or other kind of asset for your website, you can do that in Canva the way that you would in a

Vercell or Lovable, just using text to generate apps. And then that all feeds into your existing websites and Canva and things that you've built. So yeah, I mean, they're just kind of exploding still, 4 billion ARR. think people obviously, Figma went public and was a really a big IPO even if shares went down afterwards, but Canva is significantly bigger than Figma.

and has a much broader kind of audience, think, which is virtually, you know, anyone who's not maybe an engineer in a company is potentially a Canva user. So yeah, definitely an interesting one founded in Australia.

Speaker 2 (48:57)

Awesome. So if Canva was public today, what do you think you'd be trading at based on multiples versus figma, et cetera, et cetera.

Speaker 1 (49:06)

Yeah, that's a great one. think let's say 20X, then I think, you know, that probably puts them at $80 billion market cap. I think Sigma right now is at about 25 billion. So, you know, it could be as high. Yeah.

Speaker 2 (49:22)

Is it fair to say that you think it's undervalued at 42?

Speaker 1 (49:27)

42. I would think so. I think it's under. Yeah, I think it's undervalued and sort of underestimated.

Speaker 2 (49:34)

And what's your green, yellow, red? Bye, so hold.

Speaker 1 (49:39)

Yeah, I like Canva. I'm not as absolutely crazy about it as I am some of these other companies, but I think it's a really, it seems like a really well-run business. It's a great traditional SaaS business that I would love to own, but yeah.

Speaker 2 (49:57)

I'm gonna put words in your mouth. think you said you like number one, Anthropic number two, Anduril number three, Databricks number four, Canva and number five, Groq Is that right?

Speaker 1 (50:06)

You

know me so well. You know me so well.

Speaker 2 (50:08)

Not only do I know you, but I 100 % agree with you. So you don't even need to hear my opinion. I agree with Jan-Erik We have different haircuts, but we have the same brain. All right. So we have a nice compliment that we had a really good discussion. They want more of these discussions, maybe four per hour. We have another person saying, how do I really do diligence on these companies? It's kind of like throwing darts because I don't have exposures in the management team.

You need to listen to more of these. need to do your own research. You need to do research, research, research. John-Eric has a great company that gets you exposure and information on this. We have a great company that gives you exposure and information on this. So that's really what you should be doing. Feel free to add more questions if anybody has them. With that said, let's move on to the next slide.

And then disclaimer, so nothing in this presentation should be construed as an offer to buy or sell securities or solicitations offered to buy them. So all investments involve risk and possibly loss, including loss of principal, neither past performance nor forward looking information is a guarantee of future results. Next slide. Nice, we got another compliment. Excellent webinar. Thank you, Hania. You're the best.

Today's event is sponsored by the Fundrise Innovation Fund, like we mentioned. The fund offers exposure to over 200 million venture opportunities across some of the most exciting industries. So just visit fundrise.com slash Vincent to learn more about the innovation funds, portfolio companies, and flexible investment minimums. Investing in shares is speculative and involves substantial risk. You should purchase shares of this closed and tender offer fund only if you can afford a complete loss of your investment.

Funds portfolio is concentrated in technology related securities, which may carry great risk than more than diversified portfolio. Technology companies are subject to risks like rapid innovation cycles, product obsolescence and intense competition. Investors should consider the investment objectives, risks and charges and expenses of the fund carefully before investing. Information regarding such considerations, including their perspectives may be obtained by visiting fundrise.com slash innovation. Investors should read the perspectives carefully before investing. All right.

So our people have good eyes and they say, what happens at RAMP? Well, RAMP will be at a next session in the future. Thank you everybody for all the compliments. Jan-Erik, any final questions or thoughts?

What do you got to say, Jan-Erik

Speaker 1 (52:23)

Yeah, ramp is a teaser. It's a teaser. stay tuned. ⁓

Speaker 2 (52:27)

kind

of like one of the Avenger movies. It's gonna be part of the episode. right, we'll talk again soon. We'll do another five for you, maybe four. All right, have a good rest of your day. Bye.

‍