Why Data Is AI's Biggest Problem, Solana's ETF Ambitions & Your AI Shopping Agent | Aug 21, 2025

0:13Welcome, everyone, to the Informations TI TV. My name is Akash Pastrucha. It is Thursday, August 21st, and we are taking another tour of the technology sector today, folks. We are talking about fintech, about AI, and about crypto. We've got Lauren Kolodny from Accru Capital coming on the show in just a few minutes. We're also going to take a visit into Cryptoland, talking about the momentum that Solana has been seeing lately. But I want to start by talking about the business of data. Data has always been a big business in the technology sector, and AI has very much poured gasoline on the need for great data.

0:49Protégé is one startup that is capitalizing on that. The company raised$25 million earlier this month, and I want to bring on CEO Bobby Samuels and also Nikhil Basutribevi, whose firm Footwork led the funding round recently. Bobby and Nikhil, welcome to TITV. It's great to have you both. Great to see you, and yeah, thanks so much for having us. So look, Bobby, the way I understand your business is, I feel like, and correct me if I'm wrong, feels like it's like a data broker between the data licensing companies, the content companies, AI companies. Can you just help us understand the business model broadly?

1:27Yeah, so we're building a platform for AI training data and enabling folks who have rich data assets to be able to share that in a compliant, privacy-centric way with folks who are building AI models. So that's sort of how we conceptualize the business and work in a number of different verticals within the data world. So you've got like these content companies on one side, whether it be news publishers or sports leagues that have video, and then you've got the AI companies on the other side, and you're sort of this middle company that helps strike the deals in the middle? Is that the idea? Exactly.

2:04Okay. And so do you have technology in here or is it very much just relationships like I connect this company to that company? It's a combination. So we have about 150 data partners at this point across different parts of the data world. We work in healthcare, we work in media. So a lot of the killer features, so to speak, is the data liquidity that we've built. The technology is also a significant component of what we do because the raw data is often critical, but it needs to have additional work on top of that. In video, for instance, we've seen the desire for large amounts of content be consistent, but then there's also desire for smaller bits of content that have differential characteristics like the wind blowing through trees or things like that.

3:01And so we've built the technology and worked with partners to enable that sort of ultimate data deliverable to be the case. And we think that is going to continue to accelerate. Got it. And tell us some of the content companies and then the AI companies you're working with, just so we can sort of make this a really tactical example. Yeah. So on the AI builder side, a lot of them ask not to be named publicly. Everyone from the large foundation models, all the way down to pre-revenue startups where data is critical for them to get going. And it relates to our thesis that the biggest bottleneck for AI is training data.

3:49We think that's true across sectors. We think that's true across companies regardless of different sizes. And we work with folks regardless of where they are in that spectrum. Right. Nikhil, I think based on our reporting, this was the biggest check that Footwork has ever written into a company. Is that right? That's right. And so tell us about what surprised you about the space as you were due-diligencing this category of AI. What stood out to you? I think what stands out is that, as Bobby said, data is one of the critical bottlenecks in AI at the moment. And when you look at the other bottlenecks, energy, compute, models, there are enormous companies in those areas uh you know uh but in data aside from the companies that already sit on a ton of data themselves and have used that data to build their own models there isn't a clearinghouse and exchange someone in the middle who has really brought data providers and folks who want data need data together um and so i really love companies that are actually simple and easy to understand, but very powerful and have the chance to be very large.

5:06And ProJ struck us as exactly in that vein. So, Bobby, I want to ask you a question. We've written about sort of the structure of some of these deals between content companies, AI companies here at The Information. We've talked about some of these deals being flat fee deals, and then we've talked about some of these content companies wanting them to be usage-based pricing deals. What kind of deals are you striking right now and where do you see these deals going in the future? One of the things we've seen, I think across the board, is it's a new enough market that you do see different models emerging and I'd say the market coalescing.

5:47Typically what we've seen is flat fee pricing where the amount of spend scales with the amount of data that is ultimately purchased. But I think that could change over time. As technology gets more sophisticated to understand what inputs are going into models and generating and driving towards what output, I do think that model economically could change. We're so early in this world of data AI that we are continuing to see these things develop. And one of the exciting opportunities for us is helping to drive that development. Nikhil, how do you think about this? I mean, do you think usage-based pricing might be a better way to go for the content companies?

6:41I think one of the key insights that Bobby and the team at Protégé has is that different verticals actually may need different pricing and different structures. And so Protege actually started in the healthcare vertical. And especially given Bobby and his co-founder Travis's background from companies like DataVant, which is in healthcare data, that was a terrific vertical for them to start in. And then they've expanded to the media vertical and now to audio and motion capture as well. And I think, and Bobby can weigh in more on this because he obviously has more context than I have. But in each of those verticals, They work with different data providers and there may be nuances in pricing based on the type of data being licensed.

7:29But what's really exciting about Protégé is they actually have many of the same customers licensing data across these verticals. And so what you're seeing is actually customers' willingness to tweak and have different types of deals right now across different verticals as everyone is just learning and realizing the need for a massive amount of data. Bobby, say more about that.

7:56One of the things we have seen, to Nigel's point, is there's just a real demand for data across the board. Right. And in this training data world that is still fairly nascent, folks are looking for guidance, both on where do I source the data that I need, but also help me understand what's going to move the needle for my models. What does data quality look like? There are all these questions that we are well positioned to answer, and I think have done a really solid job of answering them to date. And so we really have seen folks who start with working with us in one vertical and say, hey, this has been really successful and helpful.

8:42Can you help us over here? And can you help us over here? And I think that natural growth is part of why we pursued this multi-vertical thesis is because you think that there's an opportunity for one player to really be the go-to here. And we obviously aspire to be that company. Who do you think has the leverage in this relationship between the model companies and then the data providers? We see these lawsuits between the media companies and the tech companies playing out, right? I mean, they both kind of need each other at the end of the day. Who do you think holds the power right now, Bobby?

9:21There's not one clear party that holds the power. I think it always varies. And to your point, there is this relationship where both parties need each other. Where if we focus on media, companies are always looking for net new sources of revenue, but care deeply about ensuring that their rights are protected. And that's really how we start things. Anytime we're thinking about a new vertical, we start with the question of what are the data partners and content holders that we're going to work with? Right. And coming at it from a very content and partner-centric perspective. Ultimately, for the business to flourish, we need to be trusted, not just by folks who are building the models, but also by folks who are trusting us with their crown jewels in a lot of cases, their data, and know that we will be thoughtful stewards of them and put their interests first.

10:19Nikhil, I want to come to you for the last question here. You know, another topic that we've written about at The Information is this idea of corporate data wars and some of these big tech companies like Salesforce, for example, blocking access for the Slack data to be accessed by the AI agent companies and stuff like that. What's your view on that? I mean, how do you think that's going to play out? I mean, I think, again, this is going to be one of the key battlegrounds and questions of the next several years. At the end of the day, I think the companies that build the most compelling end user experiences, the ones that end up driving enterprise growth or that, you know, consumers fall in love with.

11:05I think those are the companies that will hold the most leverage, you know, five, 10 years from now. Because, you know, they've actually built sustainable businesses. And then the question is, you know, do they build them on the back of, data that is acquired in the right way, such as leveraging marketplaces, exchanges like Protégé? Or do they have to do something else to get to that place? I think that's going to be one of the very interesting questions. Great. Well, Bobby and Nikhil, thank you so much for joining us. We appreciate it. And as more of these deals shake out, I'm excited to bring you both on to talk about it.

11:46That is Bobby from Protégé and Nikhil from Footwork. Bitcoin and Ethereum have become big household names, not just for crypto investors, but also for stock market investors who have found ways to expose themselves to the assets through ETFs. Now, Solana is the cryptocurrency that is next in line. And I want to bring on my colleague, Uechi Yang, who wrote a story this morning about how the token is becoming a lot more mainstream. Uechi, welcome back to the show. It's great to have you. Hey, Akash. Thanks for having me back. So, okay, we're talking about Solana today. Why should we care about Solana right now?

12:24So the story of Solana in 2025 is that Solana is trying to grow up. For the longest time, Solana is seen as this younger blockchain than Ethereum, which is the most established, most widely used blockchain, and is kind of known as the most favorite blockchain for minguan traders but this year we're starting to see sonana making progress signing up big name partnerships with traditional finance firms and it's also trying to build out its influence in washington dc um and we know that there's this one important upcoming milestone for sonana that's coming up which is going to be the approval of etf in the us um so once that happens uh which the market is widely expecting to be able to achieve by mid-October, it will put Solana on par with the status of Bitcoin and ETH.

13:21So I want to talk more about that, but very, very quickly, you mentioned it was the basis of a lot of these meme coins. Any popular meme coins that we have heard of that actually have Solana in the background? Oh, I'm sure you've heard of the meme coin launched by President Donald Trump. So that was launched on Solana. Okay. Yes, and there are other ones with names like Farpoint that are actually wildly popular that are also launched on Solana. Okay, so basically it's being endorsed by a lot of the big names here, certainly as it relates to meme coins. I want to just go back in time a little bit.

13:55You talked a little bit about Solana growing up. What happened to Solana after FTX, and how did that set the stage for how it needed to grow up? So Solana suffered a really serious setback after the collapse of FTX. And that's really because of the association that people have between the Solana blockchain and the ecosystem with FTX founder Sam Bankman-Free, who kind of made a name for himself as a big Solana supporter and Solana backer. So at that time, a lot of the developers abandoned the Solana blockchain. some of the startups that built on the Sonata blockchain were storing their own company fund with FTX.

14:41So when FTX collapsed, they also lost their funds. So that was a setback for the ecosystem. But things started to change late 2023, and especially in last year when Mincoin became a phenomenon And Sonata really made its comeback on the back of the meme coin frenzy because it attracted a lot of users and developers as well to trade and build on Sonata. So, okay, so we have the stage set here. I mean, look, it was, you know, closely affiliated with FTX. It's, you know, the basis of a lot of these meme coins. And yet, as per your story, it's growing up. I mean, this seems to be the next big token that is going to be headed towards an ETF.

15:27How is it changing its perception then and gaining acceptance in Washington? So in Washington, Solana this year launched Solana Policy Institute, which advocates for the interest of Solana with regulators and policymakers. I looked up campaign donations that set a disclosure and saw that Sonana Policy Institute also has made a lot of political donations to some of the super PACs across Republican parties and the Democrat Party. And Solana is also making an effort to get into Wall Street. So this year they made an investment into R3, which is a longtime developer of blockchain specifically for banks.

16:16That provides Solana a pathway into making these connections with banks and also hopefully help them to be able to achieve their goal to build what they called Internet Capital Market, which is to bring assets like stocks and bonds onto blockchain so that more people can trade them on the blockchain. And is there a reason that Solana specifically, among all the blockchains, has gotten so popular with respect to tokenizing these stocks and assets? It's the same reason with why Solana is popular with Minkcoin. It is faster and cheaper compared to some of the other more established ones. And I think one reason in particular is that there's this company called Backed Finance that is quickly emerging as a fast-growing tokenized stock provider.

17:11And they this year launched tokenized stocks such as crypto tokens that represent Tesla, Apple shares, Circle shares on the Sonana blockchain as well as a few other blockchains. But early data showed that users are gravitating to the Solana blockchain to trade these tokenized form of stocks. That's another sign that Solana is gaining some ground in its push into tokenized assets. Right. So let's take a step back here. I mean, as you watch Solana and other blockchains from here, what sort of things are you watching for? It might get its ETF, right? It might gain some more traction in the tokenizing stocks.

17:54world. It might be the basis of more meme coins, but what are you watching for in terms of how it's going to grow and whether or not it actually grows up per se? Yeah, I think stable coin activities on Solana will be a key metric to be watching for. Currently, Solana is not the most popular blockchain for stable coins. Tron and Ethereum blockchains are pretty dominant in that sense. So Solana is kind of playing catch up here, But it does have big name partnerships signed up with Visa, with Pfizer. Pfizer is a major technology provider for banks. And this year, it announced that it's going to launch a stablecoin for banks on the Sonata blockchain.

18:39Partnerships are still early stage. They're just being announced this year. So we're still trying to see what kind of traction they will gain once these products go live. Great. Well, it's a fast-moving story, and it was fascinating. We'll make sure to link it in the show notes. Thank you for coming on. That is Yueqi Yang, who covers crypto, for the information. Okay. Well, we haven't talked too much about fintech on this show yet, but innovating financial services with technology is one of the hottest spaces for venture capital at Startup Activity. And I want to bring on Lauren Kolodny, the founding partner at Accru Capital, to talk more about what she's seeing in that sector in this moment.

19:21Lauren, welcome to TITV. It's great to have you here. Hey, Akash. Thanks for having me. Great to be here. So look, fintech has kind of, it goes through waves. There's ups, there's downs, there's a lot of drama. I just want to take a step back here and just start at the outset. I mean, where are we today in the fintech story of startups in Silicon Valley? Yes. Well, as you know, we went through a frothy period, no question. But we're seeing a real reemergence right now. And I think one of the things that I'm particularly excited about, while other VCs may be hating on it, is consumer fintech. I think that we are just now seeing kind of this movement into the digitization of financial services.

20:06And what I mean by that is truly the service layer. When you think about the last generation of major consumer and B2B2C fintechs that came about after the global financial crisis, they were really bringing existing financial products online. And now with AI as the unlock and a bunch of demographic shifts underway, like Gen Z coming of age and a$90 trillion wealth transfer upon us, I think that there is a huge opening to really democratize the core services that surround these financial products that were historically limited to a higher income, more elite population. Got it. Well, I want to dive into some of these new products in a second, but I just want to rewind 30 seconds.

20:50You said some other venture capitalists are hating on consumer fintech. What makes you say that? I think that when you consider sort of the fintechs that came about in the Zerp era, so to speak, there are a lot of businesses that were built that really didn't have sustainable business models where gross profit margin, you know, couldn't support a business. And we're seeing a lot of those, you know, heater out at this point. And at the same time, there were amazing companies built in that area with highly sustainable business models. You know, the Stripes, the Coinbases, the Chimes of the world that are resilient in any rate environment that have business models that really work.

21:35And we're seeing those not only, you know, have incredible outcomes in some cases or just continue to grow in meaningful ways. Also, we're seeing them be incredibly front-footed on AI. So there's sort of this bifurcation that's happened in the market. And I think that some folks have been reticent to reenter, but we're starting to see real momentum. them. Well, we're going to come back to some of those companies in just a minute, but I want to talk about some of these new products that you see being launched in sort of this new wave of fintech. Talk about what some of these new products are that you're excited about and why you think we couldn't have launched them three years ago, but now is the moment for them to take off.

22:19Yeah. So there are two areas that I'll use as examples where we're leaning in. I'll start with kind of the automation of financial services itself. So when you think about it, it's really the service layer is now automatable. You know, there used to be sort of a lack of access to the people who had the knowledge and could help execute on complex financial processes for most Americans. And now we're seeing that change. So I'll give you an example. We recently led this year he's in a company called Alex that automates estate settlement. And maybe just for a second, no one wants to think about this on midday on a Thursday, but imagine yourself in kind of one of the hardest emotional states of your life, having just lost most likely a parent, and then in parallel being faced with one of the most bureaucratic, financially complex processes that you've ever faced, never having faced it before.

23:15And that is settling in a state. That's exactly what happened to Alexandra Mysore, the CEO, when she spent almost 900 hours settling in a state for a loved one. And through that process, she saw the opportunity to really simplify, automate, and standardize this process to really help these families that are facing grief. So Alex automates all of the admin, discovery of assets, getting access to the financial products, credentialing those, and actually going through the process of settling those estates. And not only is it the admin automation, but it's also bringing to bear for everyday Americans the knowledge that has historically been limited to really high-end service providers.

24:03So I think the implication for consumer fintech here is really meaningful. It's not just about this particular company, though we're very excited about it, but I think Alex is a representation of the opportunity that exists to really automate this service layer, whether it be for tax, benefit claims, and other household financial admin. And so in the back end there, is Alex connecting to different banks and financial institutions that hold these assets in the end? How does that work? Yeah, so they do connect in and they also come to understand the processes for the long tail of banks that are required, the cumbersome forms, the phone calls that need to be made, you know, to really get access to all of these disparate assets.

24:48Right. And last question here, I mean, what is the business model for them? So they take a percent of the settlement when the transaction occurs, when all assets are settled. So it's currently 1%. So it's a minor fee relative to the hours of your life, you're getting back, which frankly need to be spent most likely on, you know, many other things during that time. And I mean, let's come back to sort of, you know, comparing this to sort of what fintech was four years ago, the big difference obviously being AI, one key factor. So is the idea that a business like Alex couldn't have existed, you know, before this latest era of AI making that easier?

25:32Yeah, I mean, all of this kind of concept of automation of financial services, while people sort of, you know, had speculated or even, you know, stated that they could solve some of these problems, the truth is, and I've been looking at this space for a decade, as you know, none of this was automatable prior to the last couple of years. And so that's why we're really excited about the service layer. And similarly, we're excited about agentic commerce, which has similar dynamics. Well, let's talk about that. You wrote a long piece about agentic commerce. You call it e-commerce, I guess, which is kind of the new term.

26:12I don't know if a crew coined it or if it's a widely used term. But as it relates to e-commerce, what is that going to allow us to do when you combine agents with e-commerce? The thing that stood out to me was price negotiation, which is not something that I do on these digital platforms. So is the idea that if I have an agent and I'm shopping for clothes or something that I can negotiate or my agent can negotiate the price with H &M or something like that? I mean, how does this work? Yeah, it's a great question, Nikola. So let me take you through an example. So the idea with Agenda Commerce is that you can aggregate the discovery, the consideration and decisioning, and the transaction sort of into one place, into sort of an Agenda framework.

26:57And so I'll give you an example. Let's say that you wanted to go book a hotel for your family. And you said to your AI shopping assistant, in this case, your travel agent 2.0, right? You said, I'm looking for a hotel on the beach in Miami with a toddler-friendly beach club in close walking distance to restaurants. Well, you know, right now, it would be a very fragmented process for you to go figure out and do the comparison across hotels yourself. But in this new era, a travel agent 2.0 can parse that request. It can look at the family amenities and the walkability and compare all of those tradeoffs for you.

27:37And so I think there's opportunity there in and of itself, particularly around highly concerted purchases in the travel agent 2.0. But then where it gets really interesting is what's kind of happening under the hood? You need commerce infrastructure. Think like Shopify for Agenda Commerce to really provide that whole data layer and the metadata required for the AI agent to be able to provide you with that more dynamic experience. and to your question, you also need payments infrastructure. So if you think about a situation where you said, hey, I like this hotel, I want that room, but I only want you to book it if you can do so in the next two weeks at a rate that's under 500 bucks a night, right?

28:20Now, the agent has your permission to go do that and it's tracking for rates to see if the rate comes down to that level. And if it does, it has the tokenized ability, and this is all enabled by payments infrastructure under the hood, to go ahead and buy at that level for you. Where it gets really interesting is if the merchant has an agent on the other side that can actually negotiate. So now your agent says to the hotel agent, hey, Lauren's willing to book this room right now if you'll give it to her for$500 a night. and those decisions can happen dynamically agent to agent and that's where the future state gets really excited.

29:03So basically your prediction here is that e-commerce becomes a lot more of literally kind of like an auction-like marketplace where it's not just dynamic pricing but it's also dynamic bidding and this is happening on a daily basis with clothes, with groceries, with everything? Yeah. I mean, I think that's where we're going, particularly with these highly considered purchases, larger ticket items. I think there's value for both sides in making that experience more dynamic. We're a ways away from that. But when you think about that first part of the example where you set the rules as the consumer and your agent goes and executes on that on your behalf, that's on the horizon.

29:50In fact, it's happening in some cases already. And so last question for you on this topic. So in that ecosystem, then what companies are you looking at in a commerce that that really excite you in terms of their business models and ways to take advantage of that? Yeah. So I think there are really interesting companies across all three of those categories, be it the sort of verticalized AI shopping assistants themselves, especially in, again, in more highly considered purchase areas, you know, because you could make the case that more commodity products, if I just want to buy toothpaste, I'll just go to chat or the Amazon equivalent, right?