The Future of Fintech: AI and Small Business Lending in Emerging Markets

0:00Zach Marks: Greetings, everyone. I'm excited to welcome Zach Marks, co-founder and CEO at Gia. Zach, welcome to the show. Thanks so much, Ben. Happy to be here. Yeah, great to have you here. First, tell us a little bit about your background before we talk about the company. Sure. I'm from Philly. Born and raised a Philly boy. But since college, I've really been spending most of my life working in emerging markets across some fun places in Africa, Asia, and Latin America. The way I got down that world was really, I just always loved, always loved eating street food and learning languages. And so after college, I wanted to find a way to pursue that passion.

0:35So I started off my career as an English teacher. I found some English teaching fellowships that took me around Brazil, Ethiopia, and then India. And when I was in India, I became really close friends with some street vendors, some folks selling chai. I'm actually drinking a cup of chai right now. And I was trying to help these guys that are really hard workers, showed up every morning at 5 a.m. You could tell they're really pouring their heart and soul into their business just to feed their families. And these guys were trying to get a loan to open a restaurant to turn their chai stall into something a bit bigger.

1:05And I saw how difficult it's. There's no way they're going to access any formal finance. They had the local money lender, but that was about it. So as a 22-year-old, that sort of opened my eyes to the challenge that is financial access. And when I began my formal office working career, I was at McKinsey as a management consultant. And because of my interest in financial inclusion, I ended up doing a lot of work with large development organizations, the World Bank, USA, those type of folks. And that gave me my first real exposure to what we call micro finance, working with savings and credit cooperatives of farmers, working with urban street vendors, working with all sorts of small businesses.

1:43And after seeing how maybe traditional microfinance worked through that lens, I got really interested in fintech because I was curious how you could use technology to scale up solutions that would solve this access to finance challenge. And so I became one of the first employees at a company called Tala, which is really the first mobile lender. We created this really fascinating business model, which would use people's phone data to underwrite them for very small loans. Traditionally, loans require in-person underwriting or long, extensive underwriting, which would make it too expensive to serve a street vendor.

2:20and for the six years I was at Tala, we grew across countries in Africa, Asia, and Latin America, serving I think maybe 10 million micro borrowers. But what I became really interested in towards the latter part of my time at Tala was I really, I saw there had been a ton of venture capital investment in consumer solutions to access to credit, buy now, pay laters, neobanks for consumers, but really not enough to focus on small businesses. And if you look at emerging markets and what drives the economy and drives employment, it's often these small and medium-sized businesses or SMEs. And while, of course, in some ways, you'd expect that the street bender doesn't have access to credit, it would blow your mind that even a business doing maybe$25 or$30 million in annual revenue also doesn't have access to credit or the modern financial tools you take for granted in the U.S.

3:13and that's what led me and my co-founder her name is chang she was the head of product of tala to leave tala and start gf and that's perfect yeah now appreciate that background and let's talk

3:23Zach Marks: about gia and just so the listeners understand give us what's that 30 second version of gia so we understand what products and services you offer yeah we we basically build ai powered financial solutions for small businesses which broadly covers banking lending and financial operations So our core is really that we've built an underwriting engine, which can take messy, unstructured data that small businesses have and turn it into meaningful underwriting signals so that small businesses can access credit. And we have a lending operation ourselves where we use those tools ourselves. But we also, because we've had so much success and we know we're not going to be able to reach all the millions of small businesses around the world, we've now opened up that platform to other financial institutions.

4:10So now banks across Southeast Asia are using our tools to underwrite their borrowers. So there's lending. There's also banking. We deliver a great banking product, which provides better yields and better cross-border payments, just different banking products. And financial operations, which is a lot of the menial tasks of sending invoices, following up on collecting payments, paying suppliers, that sort of stuff. So we have an accounts receivable and accounts payable software, which helps these businesses to run more efficiently.

4:42Zach Marks: Okay, interesting. So focused on the emerging markets as SMBs. So when you're lending credit, what could these be? A couple hundred dollars to a million dollars? What kind of range of funding are you offering? Yeah, our typical facility is between$50 ,000 and$100 ,000. Just to put your picture of them. I'll just talk about a typical borrower in the Philippines. There's a, I just did a video with them so I can name them publicly. I want to shout out anyone who's using a lending product who doesn't want to be shouted out. But these guys are called PC Canlas and they have a warehouse where they do a ton of construction work.

5:16You walk in and it's buzzing all these laser cutters and saws and whatever, all that woodwork shop kind of stuff. And they basically make, if you walk into any supermarket, you're used to seeing the displays that hold Pantene Pro-V shampoos. Or if you walk into a mall and the little kiosk for whatever, L 'Oreal makeup. These guys build all of that stuff for Philippines retail. So their main customers are Procter & Gamble, Unilever, these big global corporations, but they have to wait three, four months to get paid and they have to pay their materials and labor costs upfront. And so while they're doing about$25 million of annual revenue, they don't actually have access to the finance they need to keep their business running and to grow it.

5:59And so we step in, we basically have a 50, I think with them, we have the$100 ,000 facility, let's say more or less, and we advance capital on their outstanding receivables and invoices.

6:09Zach Marks: Okay. Yeah. And that's a great example. And you mentioned maybe one reason that you led you to start the company was just, there's not enough focus on SMB in these emerging markets. Was there a specific experience or customer example that said, maybe we can do this better than what the other guys are doing? What's interesting is when we first started GIA, as I mentioned, what motivated me to be in this phase is probably the desire to serve the street vendors of the world. And so the first product we actually launched with was an inventory financing product focused on micro retailers. So we actually were trying to serve, think of that mom and pop grocery shop, which if you're in the US, you're probably thinking the corner store bodega in New York City, which is probably 10 times the scale of what I'm talking about in Nairobi or in Manila, just someone with really a small stall selling maybe a couple hundred dollars a day of consumer goods.

6:55And we were trying to serve them. So we tied up with the distributors who supplied those folks. And after working with a few distributors for a few months, the distributors said, hey, it's nice that you're financing our customers. We, in addition to selling to these micro retailers, we also sell to the biggest supermarkets in the Philippines. And we don't have, and could we get financing, right? Even though we're doing bigger business. And that's when it hit me. Oh, wow. This is a much bigger problem than just micro businesses don't have access to finance, but really these bigger enterprises do.

7:26And I think while that is just really a soft spot in my heart for these street vendors of the world. I think you can make a much bigger impact if you can help a small business that's employing 100 people grow to a medium-sized business employing 500 people. I think you've made a huge difference in a community and more people's lives than just helping one street vendor at a time. So that's what led us to pivot up market a bit.

7:47Zach Marks: Yeah, interesting. And you've got a lot of experience in this space from previous companies. So what do you see? Has anything changed in this market over the last couple of years that has opened up capital or financial operations for these companies? I think AI as a technology has been really great for expanding access to credit. Basically, the main thing it can do is it turns, I think, data that was previously unusable or too difficult to parse at scale, right? Because you needed to employ literal people to read hundreds of pages of bank statements and try to find some meaning in it. And now if you can do that in seconds with technology, obviously that really helps for us.

8:30Our whole underwriting engine in a really important signal is this bank statement data, which there's just little nuances that even a human, even if you had the hours, couldn't find it. The reality is machines are better at sniffing out fraudulent documents, right? Because they can tell like a one pixel mismatch. They're better at understanding what certain transaction codes are. So maybe a person would just see, oh, that looks like a thousand pesos left the account, but I didn't remember that the thousand pesos left the account every single month, which the AI can tell, okay, that's probably like a monthly loan repayment and that needs something about that business.

9:05So that's just one example of how technology is really changing the space.

9:11Zach Marks: Yeah, interesting. And with, I assume, thousands of SMBs in emerging markets, what's working for you right now in your go-to-market motion to find prospects? It's a really great question. I think the reality is in emerging markets, small businesses are still so relational. We think about this as like a digital world where everything's connected. And I'm sure everyone listened to this podcast, probably being bombarded with automated sales campaigns in their LinkedIn inbox. It just can get really exhausting. The reality is most small businesses in the markets we operate still operate based on local trusted networks.

9:45And so the best source for us to get good businesses is through our local business development associate networks. That could mean an accountant or a lawyer who operates serving the chamber of commerce, a remote town, and they refer their businesses to us. It could mean even with one of our borrowers becoming sort of an anchor node to refer other businesses they work with. That's probably been the most successful, even though it's not a very sexy answer. Obviously, we try to the extent we can to modernize that go-to-market. We have a really pretty seamless referral program, which ideally makes it so that if one customer is already our Gia customer, automatically their suppliers and their customers can quickly sign up to be Gia customers as well.

10:28But yeah, no amount of digital advertising budget is ever going to be as effective, at least today, as the in-person trust stuff. I think it'll probably start to change. The reality is more businesses are online than ever were. Yeah, and one, and I have to share about that, which I think is so cool, is one of the marketing tools we use basically scrapes Google Maps. And we'll find, you want to say, find all the logistics businesses in Cebu, right, from Ireland and the Philippines. Even if these folks aren't on LinkedIn, they probably are on Google Maps. They probably have a phone number or an email or some way to contact that business.

11:05And so we are using things like that to take it out to people with digital footprints.

11:10Zach Marks: It sounds like you're offering a lot of different product lines. Let's talk a little bit about pricing and how are you pricing your product today? What's the pricing mechanism? Well, yeah, we distinguish. There's the SME facing bit where the main product people are paying for is loans, right? We offer the financial operations software basically for free and the banking product is for free. And it's actually even just that is a huge selling point in the Philippines where there are so many fees just tacked on every little thing. You want to send money. Oftentimes you pay just to receive money. You usually have to pay some monthly maintenance fee, a monthly account minimum.

11:45So the fact that we offer a simple account, which doesn't do that, is a real selling drop. But the way we monetize the business is through our lending product, where typically we're charging around 2 % per month for different financing products. It can be slightly more or less depending on risk profile. Then that's maybe on the SME-facing side. On the financial institution side, where we sell this software platform, it's really more like a bespoke model based on the size of the contract. Just for simplicity's sake, you could think of a standard contract being around maybe$5 ,000 a month for software tools.

12:21But obviously, if you're underwriting thousands of borrowers a month, we charge a slightly higher per transaction price than if you're just underwriting 10 borrowers. And depending on the sophistication of the tools you're using within our software suite.

12:36Zach Marks: Okay. And then when did you launch Gia? What year? In 2022 is when we first started working. And I can say a quick story is 2023 was trying that micro inventory pilot. I think 2024 reshifted up market to serve more proper small and medium side businesses. 2025, we expanded from just doing financing to doing financial operation software and banking products. And then 2026 is when we just opened our platform to other institutions. And so about, yeah, so about four years in. So it sounds like maybe a lot of pricing iterations or pricing changes as your product line developed. Would you say, how many times do you think your pricing models changed over that time?

13:14Definitely. About every month, probably in some way. Probably always tweaking it.

13:19Zach Marks: Okay. Yeah, interesting. And then it looks like you've raised some capital. How much capital have you raised to date? About$3 million? We've actually raised closer to$7 million of equity. And I used to include all of the various means of financing. Since we do have a loan book, we operate. We're probably closer to$15 million. Okay. So on the equity side, so it's raised about$7 million. Is that through, say, seed rounds for that$7 million? Yep. Exactly. Mostly from USVCs. Recently, we brought on Coinbase Ventures, the Stellar Development Foundation. These are two sort of bigger names in the blockchain world.

13:53We also have more traditional U.S. theses, the churning group, which is based here in L.A., where I am. And again, some regional investors as well. Saison Capital from Singapore is our lead in Southeast Asia.

14:04Zach Marks: Okay. And was that last round then about$3 million? Yes, that's right. Okay. And then, yeah, with that last round, really curious to dive in more there. What did that capital allow you to do now that you couldn't do before with that raise? Yeah, a lot of it is for basically platforming and opening up what we built for ourselves, right? We built this technology that powered our loan book, and we were just serving customers ourselves. And then this is really for the technology investment to make this an incredible product for financial institutions anywhere in the world, not just in the Philippines, which is our first market.

14:37Zach Marks: Okay, so a majority of that sounds like going R &D investment into the product. A lot of that, yep. Okay, okay, interesting. And then what would you say was the hardest part of raising that last round? I think fundraising often is so much about just narratives and what's hot right now. And maybe at the moment, a lot of these, he's just want to invest in whatever the sexy new AI infrastructure company is. I think a lot of investors are probably just trying to get in them the third SPV of some tertiary secondary shares of the big AI giants. And so being anything that isn't that like it's something that's had an application layer, something that's in emerging markets, it's probably less sexy.

15:14And so you got to find investors with some conviction.

15:18Zach Marks: Okay. And interesting. So tell, yeah, tell us a little bit more because yeah, a lot of capital going in, of course, into AI and AI infrastructure. So if you're not, say, going into that space, maybe you have AI in the product, but that's not what you're leading with. Yeah. Was it more that you had to find investors who really knew your market and the type of software you're offering and lending? Yeah, probably. And people who had just a general thesis about where the world of finance is moving, where you believe that more and more financial applications are going to run on chain. And so we have some blockchain-based investors.

15:49There's folks who understand that, yes, AI might eat a lot of the software world, but it's probably not going to eat the actual movement of money when it comes to, say, a lending decision. I think the reality is it's going to be a while before the banks of the world kind of give up their credit committees of humans. I think we're going to be in a world where AI powers a ton of decisions and it makes these analysts job much easier and more effective, but it's not going to fully replace them. And so I think in a regulated space like finance, I think there's just room for a lot of companies to add value.

16:23Zach Marks: And so you're about four years in right now with GIA. Is there a number or two numbers that best explains what's happening in your business right now? Yeah, probably the metric that's probably the coolest to see is when we first started and we launched, we had a default rate. defaults kind of roughly defined, but think of borrowers who were not paying within 180 days of their due date, right? A lot of them still might cure, they might end up paying off, but folks who, that was like 180 days past due. And initially that number got as high as 10%. When we started implementing our AI underwriting model, which the internal code word was Asakon, our company is GIA, we call ourselves giraffes like giraffe because we tried in kenya as well and an ossicone is the horn on the giraffe that protects it so anyway but we implemented that product the number basically has has we've seen that number that was 10 percent dropped to under two percent so that's just i mean a really meaningful impact on our business as a lender and really has what's enabled us to go and sell this as a software product because we can say hey look what it did for us imagine what it can do for you.

17:32Zach Marks: Yeah, interesting. So a 10 to 2 % default rate using that AI underwriting analysis, is that something where you had that AI expertise in-house and could just start coding that that way? Or is that something you had to look around for outside expertise? Yeah, we have a great data and AI lead named Curtis. He was with us at Tala actually, and then was at Stripe Capital. And so he's led that great development. So shout out to Curtis Wilkerson. But even just with the reality is, I think you probably do need some AI expertise if you're going to be building an AI product, but I'm sure most listeners know here, you really don't have to be a technical whiz to build AI powered products, especially if you have something that is going to be more on the consumer layer.

18:15You can build really beautiful, well-functioning products without having much AI expertise yourself.

18:21Zach Marks: And then, yeah, the 10 to 2 % drop is a big drop. Now is 2 % where you want to keep it or are you trying to, what's your objective? Yeah, I think every credit, every company can have its own strategy because it depends often on the product, the target segment you're going after. You can say, oh, I want a 0 % default rate. But the reality is to get a 0 % default rate, you're really going to be lending to a very small number of customers because customers have some risk. There are companies which would be very comfortable with a 10 % default rate because they price for that. Maybe for a lot of the consumer lenders who lend to subprime customers in the U.S., you know, if you're charging 30%, 10%, you're happy with the 10 % default rate.

18:59I think it's truly depends on what your goals are. For us, we lend to businesses and we charge them a low interest rate. So we really want to keep that number really below. We really account for up to 5%, but I think if we can keep it below three, we're in a really good place.

19:16Zach Marks: Okay. Yeah. Okay. So around that number. And then, so being four years in and having prior experience in this industry, is there something that you learned building this company that you didn't believe? or didn't know when you started it? That's an interesting question. I think I've been surprised at so many steps along this journey, even going back to my time at Tala. And sometimes when you're building for emerging markets, I've been often surprised both ways on the hypothesis. So for example, when I started at Tala, I came in with, as I mentioned, I love the diversity of cultures. And I'm like so fascinated by how in like in India, where I'd spent four years living, like there's in every state U.S.

20:05is a different language. 22 languages with more than a million. It's just a huge diverse country. And so my theory initially was, oh, we're going to have to build sort of bespoke products for each of these different places. And I remember Atala when we launched, that was my hypothesis. And I probably made our team spend a lot of time probably localizing solutions. And then the reality was one product kind of worked for everyone, more or less. And so I was genuinely surprised at how scalable technology products can be once you hit on some core user behavior. At the same time, that philosophy can be taken to its detriment.

20:44Same thing at Tala, actually, when we tried to launch the product that was working really well in East Africa, in Southeast Asia. East Africa is a very heavy mobile money usage market. This was before the pandemic when a lot of mobile money took off everywhere. But we tried to take a mobile wallet-based product that worked in Kenya to the Philippines where everyone was using cash-out centers. And when you didn't localize it enough, that actually was a huge mistake. Anyway, that's one example of how I've been surprised at the amount you need to localize and then the amount you don't need to localize.

21:15What I've been surprised at Gia is similarly how what we built that was really built initially just for this niche use case of Philippine small businesses can really now be applied to financial institutions in any country in the world. We're working on a proof of concept with financial institutions in the U.S. to serve their customers because a lot of the challenges are similar.

21:37Zach Marks: Okay, yeah, interesting. And so when we look at this market three years from now, and you, again, you have a lot of experience with Tala and now with Gia, what do you think is going to be true? How will this market look in three years? That's just not obvious today. Yeah, I think probably the extent to which AI can continue to make things more and more seamless for consumers and really increase the level of service delivery that financial institutions give. So I think an example, when I started my career at McKinsey, this fancy management consultant firm that big companies we pay millions of dollars for to get their advice.

22:14And a small business like a little hardware tool distributor is not going to be able to hire fancy consultants to come in and look at their business. But now that hardware business uses a bank, which that bank probably has access to a ton of data about that business. So today you basically, there's no reason a bank isn't just like McKinsey in a box, right? If you're a financial institution, if you have that data about a business, you have the data about the local industry, you understand trends, you can basically be saying to your customers, you deliver way more than just a couple interest points on their savings account.

22:46But you could actually be their real partner in crime, be their chief of staff, be their in-house business analyst, just off the shelf without even delivering, without even having an account rep delivering those insights. And so that's what we're trying to do with our bank account product is we can say to a customer, say you have a customer who sells to both hotels and restaurant groups. We can say, hey, it looks like your hotel channel is paying three times as fast or is growing three times as fast. Do you think you should go after them more? And here's three strategies to do it. Or we can say, hey, it looks like your payday is scheduled on Thursday and you're going to have a cash flow crunch.

23:22Here's three ways to solve it. Just having those instant insights that are tailored to what's going on in your business, that's going to make a huge difference. And I think it's going to mean businesses just run better.

23:34Zach Marks: Run better, yeah. Leveraging that data with AI, which makes a ton of sense. Let's talk a few stats and figures before we wrap up and tell listeners where they can learn more about GIA. So founded in 2022, and do you have a headquarters location then? Yeah, the corporate headquarters here is in Los Angeles. A lot of the team is in Manila. We have a couple of folks in a few other parts of the world as well. Okay. And then what's your current staff count? I guess we're approaching 25 folks on the team. Okay. 25 or so. And then anything you want to share around revenue ranges? Say ARR is in the one to two million range, but we're growing fast as we bring on more of these software clients.

24:12Zach Marks: All right. That's great. And then, yeah, appreciate the time today and sharing your journey. If listeners would like to learn more about Gio, where should we send them online? For sure. Please just email me directly. Zach, Z-A-C-H at Gia.xyz. That's Gia, J-I-A.xyz. I'm very findable on LinkedIn and I'm always open to a chat. All right. Perfect. Email Zach directly if you have any questions and we'll have the link in the show notes, but Gia.xyz. And Zach, thanks for coming on the show today. Thanks for having me, Ben. Thanks.