Cross-Border VC: Scaling AI Startups Between Asia and the U.S. (Wally Wang)

We have one successful investment in the consumer application space called Higascute. So they grow from zero to $100 million in eight months. Crazy, crazy growth video generation, super popular. I think there's on the surface there's really no moat. There's Pika, there's Runway, there's so many other competitors in the similar like valuation and hypes raised hundreds of millions today.

We're joined by Wally Wang, founding managing partner at Scale Asia Ventures and one of Business Insider's 2025 seed 100 investors. Wally has built a reputation for spotting AI and enterprise winners early, often years before the market catches on and for helping those founders scale across both US and Asian markets. He has led deals that have turned into major acquisitions and even IPOs. And his cross border approach gives him a unique perspective that most VCs simply don't have. Wally, it's great to have you on the show. The episode will spotlight Wally's track record as an early stage investor, his unique ability to predict technical trends and how Scale Asia Ventures has consistently delivered early exits through strategic insights and relationships with major acquirers. We'll also explore the firm's focus on helping founders bridge U.S. and Asian markets, both for U.S. companies expanding east and Asian founders establishing a presence in the United States. So welcome to the show.

Thank you. Thank you John. Thank you Peter for having me.

Yeah, it's joining us from Palo Alto.

Oh yes, that's where I'm based. We also have team members in Seoul and Tokyo, Japan.

Beautiful cities. Before we dive into the show, do you want us to give the listeners an overview of your background?

Of course.

So I'm actually a machine learning AI scientist by training.

I came here to the US to pursue my PhD at NYU and also research at Carnegie Mellon, also at Microsoft before I decided to jump to the startup world.

So I came to Silicon Valley in.

2014, joined a Y Combinator startup called pebble that was one of the first smartwatch company. So I built the whole data platform and also the machine learning stack for that startup and we eventually got bought by Fitbit. I also spent a little bit time in a similar space, a company called Misfit that also got acquired and then moved on to the enterprise B2B market build a company called DataVisor. We raised over $150 million and VC money and we scale the businesses. We sell enterprise fraud detection and AML for banks and financial institutions as well as fintechs globally. So I built the product and also the go to market and then moved on to the VC world since 2019. So before I started my own fund I helped a large Asia conglomerate I was heading their corporate venture to invest in AI and deep tech for a bit and also advised a large family offices based in New York to look at early to later stage investment opportunities in AI. And I started the firm about four years ago and so far we have backed over 80 portfolios. So as the name tells Scale Asia Ventures we help to bridge the US with the APAC market. So our LPs are large Asia conglomerates as well as US based family offices and we invest precede seed stage startups before they got popular hopefully and then we helped them raise additional follow on fundings but also helping them to go build, go to market, expand globally and most of the founders are actually first time immigrants I mean immigrants for first time founder based here in. I mean moving here in the in the US and we can talk more about that later.

So so far we have quite a.

Few successful exits so I'm sure we will touch base a little bit on that and you know very excited to be talking about various things with you guys.

Sounds like a storied background. It's funny, I remember one of the very first things I backed on Kickstarter was a Pebble Watch I still have doesn't work anymore unfortunately I think the battery died but yeah I was a huge fan.

Yeah we have a team that are trying to bring the pebble back so there's a bunch of engineers are working on this again I think Eric is.

Also sponsoring some of that.

Our founder and yeah it was a fun journey. I was part of the Kickstarter company campaign, I mean really supporting that from the data mining and analytical part of the perspective and we did interestingly I was doing a lot of consumer, I mean customer insights and trying to understand the best marketing positions for our product. The watch collect millions of data points so we have a lot of data around the user profiles, the usages and we make predictions. We try to be helpful. Remember that was like more than 10 years ago and it was really we were trying to build a data driven company back then.

Yeah, that's super interesting. So since you've worked in this space for such a long time, you've backed some amazing companies like Hugging Face and others. How have you seen like AI shift and change and like how has kind of been watching the space for a lot longer than most VCs like giving you like some insight into how to identify winners early.

Sure, yeah. So I have been doing Android investments.

When I was an operator.

Remember I.

Went through the Y Combinator with pebble and I since started investing Y Combinator startups since 2014. So it's been over 10 years for my Android investment setup I think and when I start my fund and doing corporate ventures I also very focused on early stage. So I've been really seeing a lot of innovations over the past 10 years and know up and downs and I have always been focusing on AI technologies, the fundamental AI innovations as well as the applications. We have quite a few access as you mentioned Hugging face acquired one of our open source data labeling framework called Argelia and we also have companies like Crosstrek, the public cyber security company acquired our AI cybersecurity company called Pangaea Cyber this year. We have also two more upcoming acquisitions that are in the transaction right now. It's also large public companies acquiring our startups. I think our key observation is that there's a very good popular and healthy M and A market in the US and it's becoming even more, you know, active this year because of AI and also because of the Trump's administration is also kind of advocating a lot of those acquisitions and there's also a large organization, a large corp, they're trying to really building the next generation of products.

So from our point of view our.

Job is to really be able to spot the technology innovation at a very early on days. So when we were looking at cybersecurities we were always thinking about what's next in cyber world. Two, three years ago there were people that already started work on how to how to secure the generative AI gateway and the generative AI models. So there's a bunch of opportunities that we saw at the very early on. Hence we backed this company that this year they got acquired by Crosstrack. We have also did similar things with our other investments that I mentioned.

I would say it's quite important to.

Be able to find, to be able to predict and tell and see where the technology is evolving and try to find the best team that are doing that. Frontier Innovations.

Any insights on how to pick those teams? How to know if they're a real AI expert or just somebody that used to be an engineer and slapped AI at the front of their their tied on experience.

Yeah, that's always very hard to kind of the more you do VC the more kind of you realize that there's a lot of new patterns and unknown unknowns for us. We have been fortunate to get a very good network across academia because I came from academia in research background and we have A lot of professor advisors and top AI researchers kind of advising our firm. So we know the who's who's in the research world and also try to validate each team's background, knowing where they come from, whether they are real researchers and how many years they've been doing this versus as you mentioned, some engineers trying to pretend that they know AI or they've build AI models.

So it's quite a, so far it's.

Quite easy to tell.

But I think gradually when we kind.

Of see more people coming into this AI generative AI space and taking more time to kind of spend, I mean there will be a lot of startups that kind of educate the researchers and people that just came in this space.

I think it's gradually for us it's.

Quite important to tell the founders product sense and also understand a lot more just beyond the technology itself. So as you know, the model is already commoditized a little bit and there's the cost is also decreasing every single day, meaning that it's more and more important abuse. Build something, you know, on top of the model and you know, how, how do you, how do you define the product? How do you solve the actual problem? That becomes even more important. So that's where I think we want to spend more time with the founder to try to understand their product sense. How do they define the product, how do they define product market fit, those type of things. It's what we try to emphasize more, even more.

But of course if the founding team.

Has AI expert, that definitely will add at points, but it's not necessary.

So I'm curious to get your perspective on this. I think one of the challenges that at least we've been seeing is that it's almost like product market fit is not enough because it's so easy to build product now and there's fast followers behind you because your point, it's getting faster and cheaper every day. In fact, being a first mover might actually be at a disadvantage because you have some technical debt that a new, a new incumbent may not have. And so I'm curious, like how are you thinking about that challenge of like these companies are growing super fast in many cases, but may not have much of a moat around them. Does that moat end up being durable? Does it even matter? Because they just get enough momentum and growth that you know, they capture escape velocity. And I'm curious to hear how you and your partners are thinking about that.

Fantastic, fantastic question. We've been thinking about this for a long time and of course there's a lot of debate among the VC community about the moat. Our view is that we shall split to a consumer or B2C versus enterprise. B2B.

Yeah.

In the consumer B2C market. Yeah, you're right, there's not that much of a first mover advantages and I think in fact a lot of the there's enormous examples of the first movers back in the days like character AI cohere, things like that. And honestly I'm not super sure about perplexity but TBD on perplexity. But I would say, you know, there's still a lot of unknown and hard to predict who will be the ultimate winner. What is the moat? But at least I think one thing clear is that you have to I mean the anthropic of the world, the OpenAI of the world, who has the latest frontier model capability and keep pushing that capability powering the applications in the consumer or in the application side they have the moat. But others, you're right, you know, just a simple wrapper. You know, a lot of hypes, it's hard to tell. So our we have one successful investment in the consumer application space called Heliscube. So they grow from zero to $100 million in eight months. Wow. Crazy, crazy growth video generation app super popular. I think there's on the surface there's really no moat. There's pika, there's Runway, there's so many other companies in the similar valuation and hypes raised hundreds of millions for us, I think you need to boil down to the fundamental technology they're building. So they're building a very important infrastructure for the model routings for videos so they can pick and choose and also aggregate different varieties, hundreds of different models for different use cases, whether that's open source model, whether that's closed source models. So there's a bunch of that. And not only to increase the video quality and the video effects but also reduce the cost of generating each video. So there's some very deep technical moat that drives our investment. That's why investors are super excited TBD on whether they will be the ultimate winner. But I think they have a shock because of the underlying technology.

Having said that, I think that in.

The consumer sector it is quite hard to tell yet what exactly the moat will be because honestly the AI applications they are lacking the network effects which as you remember in the Ubers of the world BBS back in the days, there's very strong, even Facebook, there's very, very clear varieties. I mean the network effects, they have built in. So there is clearly a first mover or kind of early mover advantages. But in our AI generated AI applications we have not seen that moat yet. So on the consumer side, on the.

Enterprise side, I think it's quite easy to.

I think we're still very strongly believe in the very a typical B2B playbook which is you have to build a very verticalized end to end solution to your specific customers. So that's why we backed some of the leading legal AI agent startups. One company called Solve Intelligence, they're building AI agents for legal, patent, legal and illegal for you to file provisional patents or for legal firms to review and, and those things.

So that's a big enough industry and.

They have built a kind of a first mover advantages but more importantly start to build their own data mold because they work very deeply in that vertical with the top lawyers for years to accumulate that data and model moat as well as their reputation in that industry. It's very hard for newcomers to kind of claim that they have a better model or they, you know, no one will trust you. And I still have millions of good cases that you already filed. So that's a very easy to understand example. We have also look at for example physical AI space. It's also quite important to build an industry focused end to end solution and that is the stronghold because the largest customer we were looking at this deal that has FedEx is their larger customers. So FedEx will not pick randomly a small startup that just started maybe a month ago or six months ago with the latest foundation model to try and pay them millions. It will take years for them to put that in the real production environment and doing all those integrations, etc. So that's the mode that integration, the trust is the mode. We also invested in a company called Cresta which is one of the fastest growing AI agent in the enterprise in customer support and cost center space. We're growing very, very fast. I wouldn't really reveal their actual number, but growing on par or even larger than Decagon and similar to Sierra. So they are. One of the things that we really like is that there is a very, very competitive market. A lot of incumbents, a lot of newcomers, a lot of large, very well funded startups, AI native startups like Sierra and decagon. The reason why we think Cresta has a place in this big market is.

That.

They integrate very deeply with the actual call center systems in the large enterprises. So imagine you have 10,000 people call center facilities, there's a lot of hardware that are powered by Genesis, Avaya, all those you have to plug into those and also you have to work in the different desktop environment, different softwares that the call center operators are using. So you do that. It's not just about the model but also about the integrations to the workflows that operators are using day in, day out. So there's a lot of those technical nuances and customizations or type of like integrations they have to do that is that's creating the mold and people build, I mean the team build that for. For years, really a few years. So the model itself is commodity but, but everything above the model becomes quite very important especially for enterprise. Yeah.

What are your thoughts on. I heard recently that you know, one of the big challenges for a lot of these businesses like Sierra for example was that growing really fast customer, you know, clearly providing a lot of value. But there is this big question around how much longer will it be before customers just build it themselves. And so some of the pushback has been that companies are not able to secure the long term SaaS contracts that they've been able to in the past because their customers are saying I don't know, know what's going to be the top performer next year, let alone four, three to five years out. And so it feels like there's a lot more push towards like usage based pricing and a move away from SaaS based. I'm just curious what your thoughts are around that if companies will ultimately be able to use AI tools to build their own solutions or if you think that's, that's just unrealistic and, and they're just being a little too conservative.

Yeah, I think that's a very common skeptical and I think that's build versus buy is of course one of the key criteria or we have to underwrite and understand about each businesses for us. I think the one more interesting question is also about what if Anthropic tomorrow come into this vertical.

Right.

What if anthropic becomes because Anthropic already launched their coding agent. What if they got a legal copilot agent or you know, very deep dive into customer support or whatever.

Yeah, whatever.

So our view is that you know, we want to look into those again vertically integrated end to end solution. So if it's a very horizontal solution is a little bit could be a next step for a foundation model company to develop Anthropic. But if you are very vertically integrated to a specific industry, for example in the legal industry or in the healthcare industry because of the Entry barrier requires you to target a completely new category of customer profiles. Like for example in a bridge, which is a very popular healthcare transcribing agent for healthcare practitioners is really. You have to sell to the clinics and the family doctors and the specialists. So that's not a trivial business to sell to. For a horizontal anthropic type of company to sell and therapy can sell to other technology companies, not a not a big issue. But if they want to sell to the millions of the small clinics or the hospital family doctors and specialists, that will take long time and also extra efforts. Plus you have to do the integrations with the EPIC with all the system of record in that specific industry. So for specific industry focus the vertical agents, we think they're safe, they're fine. So that we wouldn't see much of a. Of course there's no way to DIY because the hospital, the family doctors would not be able to do that themselves. Or the EPIC of the world doesn't have any AI capabilities. So I think there's for vertically focused AI native agents we think it's safe and it should be a pretty safe bet for companies like Sierra, like Cresta building on more of a horizontal across industry. A call center, contact center or marketing agents or there's a lot of video, video or market video ads, agents, etc. Or even like Geo like Generative Engine Optimizer or optimization agents for enterprises to boost your brand and your marketing and also all those, you know, sales agents. I think that's, that's something that I think the foundation model, when the model capability increase there's pretty easy to either let the customer DIY or let anthropic to kind of like focus on that and launch a product. So that's more of a. Could be a concern especially for plus like all the consulting companies like the Accentures of the world, the McKinsey of the world activity and trying to build customized solutions and agents for each of the large enterprises. So it's quite easy to kind of pick the latest model and then to package that and solve specific, you know, building that horizontal like Chatbot or the, you know, the sales agent. So in our view we love the, I mean horizontal agent companies who sell to the largest company in the world because we feel the largest company, the big enterprises they need a specialized solution versus a DIY versus a more of off the shelf solution with some customizations from the consultant firms. Our view is that you need a very dedicated large enterprise solution to solve and deploy and making sure it's secure and Making sure we have the 247 support if anything goes wrong and it's the team of the experts that is so that's why we believe in Cresta for example all of their customers are large Enterprises, Fortune 100. Sure. We rarely don't like the mid market or the small medium sized businesses. I think that might be easier for what you just mentioned. DIY build themselves right Use off the shelf solution from the model company plus some consulting services.

Yeah, no it makes sense. Yeah it'll be interesting to see how it ultimately plays out. I mean one of the things that's challenging to know is the technology just keeps improving so fast that it's. I don't know. I've never felt like it's been harder to predict what the next five years will look like than it is right now because of how dynamic the environment is. But yeah.

Yes.

So I want to switch gears a little bit and talk a little bit about your cross border approach and would love maybe maybe we start with just. I would love to understand better kind of your thesis around how you think about cross border investing in founders, bringing them to the US bringing companies to Asia Asian markets why you think that that's a really interesting opportunity and I assume is a critical part to helping you win deals.

Yes, you're absolutely right.

That's the whole purpose of having a.

Unique positioning in the market because we are also competing with other the hundreds.

Of different VCs out there and yeah.

I think practically speaking is quite a unique advantages and we strongly believe that.

Because we invest in a lot of core infrastructure you know deep tech infrastructure level products. So for example companies like vast data providing the high performance storage SSD for the AI compute so they are you know $30 billion private company hasn't public yet but they're kind of in this we really like those with very fundamental technology innovations and those companies when they at certain scale they have to kind of expand globally and because it's such a horizontal critical infrastructure play so it's very easy to translate and sell to other market and in fact quite a few of our LPs limited partners our investor are the largest system integrators and resellers in Korea Japan such collaborations they can collaborate with our portfolios is very important for them and in fact they are helping last data they're helping Another example is weaviate our vector database company and another example is CAS AI they do the GPU and CPU cost reduction for your cloud. So there's a bunch of those our US portfolios that are already working with our strategic LPs for those reasons and hence they will sell to the Samsungs of the world, SKS, LGs and KDDIs and Mizuho bank in Japan, et cetera, et cetera. And it is very important to understand that because of a language barrier and the cultural barrier, it's better to work with a system integrators or a channel partners to help the startups to expand. Otherwise it's very hard. I think we'll start really well. By the way, Japan is the second largest software market in the world. So the domestic Japanese enterprise market for purchasing US software is gigantic. And there's enormous opportunities in the financial services industries, et cetera. So our LPs by putting into money into our fund will get access and working together with our portfolios. And our portfolios will also benefit a lot on the application side of the world. That's where we invest in Cresta. We invested also some physical AI, physical intelligence applications in logistics, space and warehouse, et cetera, et cetera.

So there's a bunch of more of.

A cutting edge innovations that we see. The US is really leading the world and we don't see anything that is happening locally in Japan or Korea yet. But you need a little bit more localization if you pick to go to those markets because of the again there's more language and services oriented because you are touching the actual business problem. You're studying application software, the targeting talking about you need to integrate to the business workflows and integrating to the employees, have to manage that, et cetera, et cetera. So there's a lot of that, not that many of our portfolios are ready, especially if they're building an application. So our strategy is that if companies like Cresta, companies like some camera VRM companies in all this space, once they're ready, then we start doing introduced and helped. If they're not, it's totally fine. If they only focus on the domestic US market or English speaking market, we feel that's a little bit more. It's harder to go to a foreign country than the infrastructure level startups. So our way of, you know, our way of thinking is that we will help maybe find some opportunities in locally in Japan or locally in Korea that are kind of learning the playbook that the US is building in the application and trying to build a localized AI native applications. So one simple example is for example Cresta or the call center, contact center space. We met a few teams in Japan that are trying to localize building a Japanese version of those companies. It's a big market and we believe that there should be a local player that, that is like the Sierra in Japan. And in fact there's a bunch of companies that raised decent capitals already in Japan. And looking, I mean looking from that perspective, it's like you are basically copying from the US to Japan and there's still like three to five years back in behind what we see in the market in the US So I think that's where we want to spend maybe a little bit more time in the future to help maybe incubate some companies locally there to learn from the U.S. the other way, the other direction, which is helping Asian founders to scale globally. That's another thing that we very specifically focus on one single sector for the moment, which is the robotics and hardware space. So we believe that East Asia has very strong hardware capabilities that they can supply, you know, even have some surplus in the, you know, the human noise or all the AI robots they kind of produce or they have manufactured. And I think there's enormous of opportunities either to consumer or to enterprise here in the US and what we want to do is that we will take advantage of those, you know, founders or those experts or those suppliers from East Asia and build a US based enterprise or consumer robotics or AI device company. So that's something that we've been really actively working on. You can see that maybe incubation or kind of like really trying to build that businesses and it's very specific, very, very selective of course like to, to, you know, to, to, to really we spend a lot of time with co building that businesses. So that's the, the other direction from asia to the U.S. i think we.

We want to start from there first.

Yeah. So there's a certain areas that as.

A, as a cross border VC firm.

We can play and that's where we, we find it's extremely interesting, especially at the current moment. Yeah, yeah.

Oh, sounds super interesting. What, what are some of the challenges that you found or you know, what businesses tend to be the most successful at succeeding at cross border?

Yeah, I think the, in terms of the. You're talking about the succeeding criteria.

Yeah, like I'm just curious, like what are some of the insights that you've discovered over the years of working with businesses that are coming either direction and you know, what types of businesses or approaches or founder backgrounds tend to be the most successful and which ones tend to really struggle?

That's a great question. I think there's a couple of things for enterprise. I think we have seen a Lot of great examples in the more of developer tools or infrastructure layer that are.

Universal and people can scale. A US business can scale very easily to a global scale. We are talking about for example Splunk, we're talking about.

Elastic Microsoft snowflakes, databricks, the world, et cetera, et cetera. So that's why I think the playbook that I've talked about for our infrastructure portfolio is quite easy that can scale.

Following the same playbook.

And even the people that you hire in Asia are the ex snowflake, their ex, you know, databricks, et cetera, et cetera.

That's how I think anthropic is hiring.

The same way and opening and in Asia. So that's very straightforward for the application side of the business like the zoom of the world or the you know, the slack Dropbox of the world. There's not that many successful I would say the small. It's quite small or the market is not as gigantic as we hoped. I think because I think.

There'S various.

Of like local competition or there's cultural barriers, et cetera, et cetera. So I think in the application layer.

Is a little bit usually it's a.

Little bit harder to sell to. We have seen some success but it's still early to tell. But I would say I would be more bullish if we can see some local companies, homegrown companies in those applications layer of businesses.

Yeah, for the consumer side of the thing.

That's something that is more interesting for us. We have I personally because of the pebble company as you as we first mentioned on the first company, Smartwatch company.

Yeah.

We try to launch in Asia and misfit we did successfully launch in India, in greater China and so we manufactured in Korea by the way at the time. I think in consumer market it's kind of more global and universal. But you have to find the Elon Musk of the world who understand the.

Market and who moves really fast and who can really collaboratively working with a.

Lot of different stakeholders in different cultural and different parties. So that's where I think not a lot of founders that that have this capability.

So.

If we take a consumer or application side of the businesses and the founder has to be very important. I mean what we care is how quickly they can learn and pick up new things and how quickly they can adapt to that's.

What advice would you have for a founder here in the US that wants to tap into Asian markets?

Yeah.

So come to talk with us. That's of course super important. And we Want to advocate that because not so many great partners here in the US that has this resource, this new resources. And I think also there's, and also the founders can also talk to other founders who has few things before or failed before.

I mean taking learning lessons. As I mentioned, there's a lot of cultural barrier and business settings barriers. You have to work with different partners and different stakeholders, those sort of things.

But I would think that it's a.

Big market, especially considering the population, the GDP growth, the willingness to purchase procure softwares, et cetera. You know, this is quite a large market.

So is it easier for founders if you have two founders, one in the US trying to go to Asia and one in Asia trying to come to the US who has the easier time?

I mean who, who is easier?

Who has the, which opportunity is it easier to jump to as an Asia founder to jump to the US market and sell or as a US founder to go to Asia?

That's, that's very interesting question. It is the, it is the Asia.

Founder come to the U.S. it's easy, it's easier.

How much of a magnitude difference would you say there is?

Maybe ten times. Wow, interesting.

Because the, honestly the, because that's, that's a cultural thing, right? I mean everywhere in the world are heavily, heavily influenced by the American culture.

And people study the U.S. i mean even if they have not been to.

The U.S. they've studied the U.S. culturally.

Economically and business wise and for a long time.

So that's why people have a basic understanding and the US follows a lot of very transparent and a lot of the very easy to work with rules.

And, and, and, and policies.

The other way around is very hard.

So yeah, if I was one of your robotic companies, let's say I'm doing like robotic operating systems, so maybe just software only. Would you advise me to jump to India or. Not India, but Asia and when, like what would be the criteria? You're like, hey John, it's time go hit it.

Yeah.

So my advice is that if you are building a software for robotics, you shall stay here. Stay here, sell to the US but import and get your suppliers from Asia.

Not to sell to. Not to sell software.

Not sell software. You are trying to sell here first and then once you are good, then you can sell over there. So that's usually what we find. That's why we invest in the US because yeah, I mean really it's our portfolios in the US because you have to get this market first. This is the largest market. And then you think about There and I think there's a big opportunity. That's why you know, we emphasize on the Asia. To us we only focus on one single sector which is leveraging the supply chain. Robotics from Asia. Build the company, build the company here in the US and there's numerous opportunities if you know like 1x or the, you know, other. There's a bunch of, not that many. There's a couple of humanoid startups here in the, in the US but they're not really, in my opinion they're not.

Really on par with what we have seen in Asia.

They've developed much more advanced in, in terms of the hardware form factors.

Yeah.

So, but so that's where we want to, there's a huge arbitrage opportunities.

That's why we want to pick some.

Asia, you know, founders or suppliers and.

Build the business over here.

So as we wrap up, I'm curious what's one bold prediction you have about the future of tech that maybe other investors don't agree with you on?

Yeah, that's a, that's a hard question.

It is a hard question.

Yeah.

Because this area has been overheated.

Yeah.

It's such in a hype and every single layer, every single sector has a lot of money flowing in already a lot of competition. So it's very hard to say anything that are not hot yet. I would say from our point of view, we think, you know, we think so we focus on large enterprises. We think there's still room to grow if you're selling to large enterprises because large enterprises will need a lot more and they can spend a lot more. And not that many startups. We haven't not seen many startups for example, targeting the largest bank of the world, the startups. Now it's just selling to each startups most of the days building agents and calling each other agents, selling to each other agent companies. We want to see the company selling to JP Morgan. We want to see companies selling to the largest, you know, the, and also the government.

Right.

And that's big budget, not that many competition. So we think that's a area for sure. We like and the. We talk about robotics, we like that we think the physical AI will come. We don't know when but it's a big market. Not that many startups working on that. Honestly in a very, in a very realistic way. I do not like the futuristic like movie style type of videos that you can shoot whether you're human humanoid robots. We want to see something really, really, you know, solves real business problem and make real roi. So that's another area I think we don't see many. And I think as I mentioned, the Asian market is lagging behind, especially in some of the call centers or the sales agents or other things. I think, for example, in Japan, I believe five to ten years you will.

Then see the anthropic, the open AIs.

Of the world in Japan that are just started to become popular or become the enterprise start to adopt.

So there's a lot of opportunities I think still lacking behind.

In those markets. Yeah, those are pretty much our predictions.

This is some good, solid predictions.

Yeah, thanks. Yeah, I try to be insightful. Yeah.

Well, this has been a super fun conversation. I feel like I've learned a lot.

About.

This whole idea of cross border investing as well as kind of digging into some of your insights around AI and the future of enterprise software and robotics and so forth. So thank you so much for joining us for this episode.

Yeah, absolutely.

My great pleasure.

We're lucky to have you.

Thank you.

Wally, where can our audience find you if they want to learn more or engage with you?

Yeah, so they can reach out to me at me on LinkedIn by searching my name and they can also send me email.

Wally.

Awesome.

Cool. Well, thanks again for being on the show.

Sure.

Thank you.

Thank you, Peter.

Thank you, John.