Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Account Aggregation: the next UPI | Kumar Srivatsan @ Fego
430 Plays2 years ago
Account aggregation is a process that enables one to gather all their financial data from multiple sources into a single location. This technology is set to change our relationship with finance fundamentally. Fego is an Open-Finance platform that is building products on the account aggregator framework.
Know about:-
- 2 types of lending models
- History of account aggregator framework
- Reasons for multiple account aggregator operators
- Finance is no longer a separate vertical but an embedded feature
Recommended
Transcript
Introduction and Tech Landscape
00:00:00
Speaker
Hi guys, I'm Smarsh Chaitan from Seego and the co-founder and CEO, Seego.ai. We're an open demand startup.
00:00:18
Speaker
In the West, the big tech companies own the platforms that run the digital rails of the economy. In India, the government is building neutral frameworks which tech companies can plug into and run the rails of the digital economy. The best known example of this so far has been the UPI. But the fintech industry is even more excited about the account aggregator framework.
Kumar Srivatsan and the Indian Fintech Scene
00:00:39
Speaker
In this episode of the Founder Thesis Podcast, your host Akshay Dutt is talking with Kumar Srivatsan, the founder of the fintech startup Faygo. Faygo is building products on the account aggregator framework and this episode is a masterclass on what the framework is and how it can fundamentally change our relationship with finance.
00:00:58
Speaker
As you listen to this conversation, keep in mind that this was recorded a few months ago. And if you like such insightful conversations with disruptive startup founders, then do subscribe to the Founder Thesis Podcast on any audio streaming app.
00:01:18
Speaker
You are a native of Chennai, you grew up here? Yeah, I'm in Bonham, Brad here. It's been almost three decades, west-space of Chennai. I did my schooling here, I did my Chardagonnitsi course here, I'm pretty sure it's the coast-space of Chennai. Of course, qualifying, I was working with Anstanyaan, Ralbhan and Scott twin, then started my own advisory practice focused on start-ups called QAD. My chance currently on it is no longer actively part of it.
00:01:42
Speaker
So that QED was like a CFO with a focus on start-ups? Yeah, it was essentially a financial advisory firm. So what made you move on? During 14-15, you had for us, you had lending club. All of these guys built new-age capabilities, digital lending capabilities, and that was pretty intriguing. And that was when we realized, why don't we build something? Why don't we try to do something like that for India? Because there are a lot of open spaces, a lot of wide spaces which were not okay.
00:02:10
Speaker
And thus, after credit was won, we were focused on certain sections of the white-coilet audience. I started that along with my brother, Shrikantan, who was an ISB grad, who was earlier the head of special projects at NetMeds, then moved to KPMC, then held up the credit for the early days, and he was part of the founding team and the co-founder as
Retail Lending Challenges and Strategies
00:02:29
Speaker
well. So the idea was, well, if you look at retail lending institutions, most of retail lending institutions used to profile employees based on the employer to which they belong.
00:02:40
Speaker
And that was a comfort factor as a proxy for salaries, repayments, etc. And that eventually became a list which was, which is pretty narrow and you've had all of these entities categorized into AMEC, government, etc, international and that excluded a lot of entities which are building capabilities, which are growing over time.
00:02:59
Speaker
And consequently, employees belonging to those uncategorized employers were denied unsecured period. While secured was never a problem, hypotected loans were never a problem. So we said, why don't we tie up, do a two-pronged strategy? Why don't we tie up with them, provide it as a financial wellness offering? Why don't we tie up also with balance sheet players? We also hold water and VFC license later on. And why don't we build our own balance sheet and also work with other balance sheet players like DNI for DMI, et cetera.
00:03:25
Speaker
We tried to build a co-lending capability, build our own risk models. We did a B2C approach and also had a B2C approach and we tied up with employers as I told you. It took a lot of iterations to find the right sweet spot because if you look at the retail lending basket, there are so many competitors across every ROI bucket. You have the typical PSBs from the initial lowest-cost ROI, you have top-notch private sector banks,
00:03:49
Speaker
You have some MBFCs and it goes up to Fullerton, you have payday lenders, India bulls, all these guys that feel high, higher with the dairoys. So there were certain suites purchased which were not catered to. That is the size of the Indian market. So you could build a car out of the pie and build capabilities. That's what we tried building and that was the initial research in the market. What was the pie that you were targeting, your sweet spot?
00:04:14
Speaker
Most of the data-ending institutions, they provided unsecured loans to employees belonging to only categorized employees, which they profiled indirectly based on their receipt.
00:04:25
Speaker
The pie that was not catered to on the retail institutions, the entities that were profiled found only 8% of the total MSME smashed employer bids. By which I don't mean mom and pop stores, but typical NTPs that are registered already have a banking relationship. So 92% of the entities which had an existing banking relationship by Raffa Kananda, Krasa, Kananda.
00:04:48
Speaker
did not, were not profiled by their own banking partnerships, by their own banking partners, and thus their employees were not provided unsecured loans.
COVID Impact and Fintech Innovations
00:04:58
Speaker
And that became a vicious cycle because until and unless you moved to an entity which was providing say a TCS or a cognizant, you wouldn't get an unsecured loan and you always had to go around with secured loans, secured loans, hyper-decorated loans. And now the market has definitely worked. But when we started six years back, that was the case.
00:05:15
Speaker
And we tried to double down on that segment and we tried building capabilities over time. There were a lot of in-texts starting at around that time, targeting, I think, patients, possessed money, India learned. All of these were looking at the same market only. But each one had a different model. What was your model?
00:05:35
Speaker
So Rajad Wadi was an early version of the land wage access that you correctly see, refined doing it, you have quite a few players doing the early wage access program right now. So we tried other than B2C approach, we tried working with employers. So we tried working with employers which had minimal stake account and there were a lot of employers that we were working with. And we bundled credit as an offering along with financial education, along with savings and with. That was the major differentiation.
00:06:01
Speaker
While this was focused on the combination of V and FL and checkout, patience was focused on relatively larger ticket size, relatively medium to not personal loans. We had a different approach to the market. It was credit first, but it was not necessarily completely sourced directly from the market. We do have a B2C approach and that helped us with.
00:06:20
Speaker
combination of relationship collections and the much more stable portfolio compared to other competitors. Got it, got it. Okay. And so why aren't you still running that business? For every Janji, there's always a point in time where you don't start with that intention. But I read somewhere, these are saying, the moment you start a venture, you're working towards a full stop. So how long do you drag it? How long do you create value? Because what was the path to come down? Essentially, that is a journey. How do you put in process? For us, it was never
00:06:49
Speaker
In general, it was essentially a macro event. We had COVID footer, had different plans, and put forward. And being an ending starter, we raised only minimal capital, where we actually hit the market just before COVID. You had the series of blacks won in it. You had ILSs, you had DHFL, you had ICIC, the sentiment was pretty bad. And we were about to be acquired by a few consumer companies. We had a deal on the table. April 17th, last year, they pulled out, citing COVID. I don't blame them for it. That was first major event.
00:07:19
Speaker
And we took a call whether we wanted to start something or join because we had offers from the same acquirers to say whether to leave credit for us. We had large kinds of internet companies that said, Mark, can you leave credit for India and India and Africa. So this consumer internet, by this you mean e-commerce or an aggregator or whatever.
00:07:38
Speaker
Travel aggregation platform, and then we also had one of the largest apps that focus on blocking spam calls that wanted to build a credit journey. And we met the point that we came down. Then that turned out into an active hair search and employment offer. But I still felt there were problems that were never solved because we had an average experience for four years. So many times reinventing the wheel, we realized what was a huge pain point that was not solved yet.
00:08:03
Speaker
And that gave us an insight view as to how to build things. You make so many errors during your first venture. When you're building the second one, those seem like so much research. And you really think, how did you even try to do something? How did you take advantage of something like that?
00:08:19
Speaker
I guess it's that learning culture that people sort of value on. Now, I realize why second-time founders are valued. Because they know what problem. You've already covered a large portion of the journey. You just want to build on a foundation where you know which mistakes weren't the word. And it's that tremendous learning. I don't think, had I been in my employment, same journey, or had I even done an MBA, that couldn't have maintained the same.
00:08:42
Speaker
learning, constant learning. In fact, the level of learning that we did was, it was the ICO era in 2017-18. We said, see, lending is relevant in Web2. Lending can also be relevant in Web3. Why don't we see if we can try to do something. You had platforms which used to take digital assets, have a fixed free-fixtality ratio based on volatility and provided liquidity to borrowers.
00:09:05
Speaker
He said, why can't we do it as a centralized? There's no point doing decentralized because pairing tokens for creating liquidity pools would be large. So why can't we do it simply? It still makes sense. And we wrote our own white papers. That was at the time of starting. Somewhere in between when we were like, product market fit was there, but you don't know whether you would scale or not scale. That's right. And that was primarily because we were building a lending platform from scratch. You're not hardcore lenders ourselves. So you always move sideways. You see what works for you, what does not work for you.
00:09:32
Speaker
Now you realize you take a problem, single problem, scale it, you focus, the IP that gets built is super valuable. Those are the learnings that we realize
Risk Assessment and Learning in Fintech
00:09:39
Speaker
all the time. So, like one of the mistakes you made at the time of Opta Credit was that focus was not laser sharp. You were possibly experimenting with technical ideas.
00:09:49
Speaker
Yeah, experimenting a lot. And that, I wouldn't say that was the problem during our scaling phase. That was a lot of options we kept on evaluating until we did, okay, let's do this, B2C, B2C, and B2C model. Until that time, it was a whole host of experimentation. So what were those scaling up mistakes that, for example, early salary does exactly the same thing, what you do. There are a lot of players who have moved on to the B2C model and working with them. But what I realized was
00:10:17
Speaker
In lending, there are few models that are 100% digital and a lot of these models are a physical element to it. Collections is the goal. At the end of the day, if you look at how the market was, then people were valued based on your dispersal run rate. It looks stupid right now.
00:10:35
Speaker
But I think finding them if you are able to hit a 2030 growth and read per month, you won't be eligible for series S. I think that is the case today also, right? Lending hacking people have opened up for other pieces of the fintech ecosystem, what you say might be right. But for lending, I think people have opened up essentially to play.
00:10:52
Speaker
See, your valuation for a lending startup has to be linked to book. If you are going to try making it link to disburse in, I think markets will always find out they're in for a disappointment because lending is, and if you look at someone who's going to give you value and disburse,
00:11:07
Speaker
your NPS will always be straight because your denominator is always going to get average. Can you break that down? NPS will be skewed because denominators is averaged out. Just explain that line. You've been dispersing 4 crores a month when suddenly the next month you're dispersing 10, 11, 12, you're scaling pretty fast. Ideally, the way we did
00:11:26
Speaker
was also adding this to the funding market. Then we built a portfolio, curated refinements behavior, identified what COVID's work, what COVID's did not work. That means in an average, say an 18 month loan period, you'll have to wait at each six, seven months to see which are early quick mortalities, which are early signals of stress in your portfolio.
00:11:46
Speaker
And when you go fast, ramp up. What happens? All of these numbers get blended in the overall dwell. Every month you add 13. There's always a pre-AL guy that gets collected. The moment you collect a pre-AL, at least for 20-30 days, you think he's good. And the next AM is 30 days away.
00:12:02
Speaker
So in a medium that I've learned, that is a problem. And after a point in time, you suddenly build a 300, 400 crore book, and your NPA's start showing up. And then you cannot put your elections policy, then you've got to add anything you want. Stress has happened. So by NPA, you mean non-performing assets, like bad debts. Yeah, non-performing assets. So could be your bucket, not necessarily a 90-day for collections, anything that is about the 30th day is bad. And accounting wise, only a 90-day is bad.
00:12:28
Speaker
But for business, anything about 30 days a day, if you are even in a study day, the cycle is bad. We realize the amount of respect that you have to give to the incumbent financial institutions, your bhagat, your mood, thoughts of the world. The level of production, finance and capability in the relationship they have is tremendous.
00:12:45
Speaker
But what they lack was something that now they're trying to build. And the combination of the tech slaver to finance is all that balance will always reside in favor of collections in a lending startup. That was a hard learning that we realized. We realized pretty early on, but I think it's almost also a little drawn to be too ahead of the market. What we did was we built a small quarterly, but curated it. But everyone was asking for disbursals. By the time you went to disbursals, we had a series of black monuments. No, I did. I did. I did.
00:13:12
Speaker
I think even now, there are multiple lending models. Even if you look at a payday model or a BNPL model or a checkout financing model, I think essentially the idea is to say it just varies. You only have a testing ratio. How many users are you going to give bet on and see how many are you going to convert into good? That ratio changes. For BNPL, the ratio might be larger. You might not spend that much money on file costs. You might not spend money on, say, fraud analytics because it's a business of quality versus quantity.
00:13:41
Speaker
average ticket size of 5000, the money that you're going to be spending on pulling a bureau or pulling other costs will be less. So, typically what people try to do is give a smaller loan, don't do all of these checks, just check for this repayment behaviour, then try to up the loan.
00:13:56
Speaker
So now the scenario is BNBL, you typically make money only in the second loan or third loan. It works if you have a funnel to upsize the loan. For example, you mentioned patience. Patience and lazy pay work wonderfully well. Lazy based BNBL, they do call out the guys who repay on time. And if they want, they can upsell the customer to pay since the medium sized loan.
00:14:19
Speaker
So, they become a quality acquisition engine, but stand-alone VNPL, you'll have to see how your economics works. So, these are our learnings that we realized for which sort of retail and product work works, what doesn't work. And I think the journey has been super fulfilling, super interesting. Though, sadly, it came to an end the previous startup, but the learnings were immense.
00:14:40
Speaker
So if I had to recap some of those, one learning you had was if you focus on portfolio quality, do you cut down your non-performing assets, then you will have to go slow on dispersal. And going slow on dispersal means that investor interest will not be there because they will see this as a slow growth startup.
00:14:57
Speaker
Typically, people say, don't build businesses for investors. I agree. I didn't build my previous business for investors, nor this business for investors. It's for my clients to make money. And if we need money for investors, and we're interested, we make a deal. But what happens in certain scenarios of flux, and what happens in certain scenarios of market heating up, is people take a bit, and those guys, the ventures who raise money,
00:15:23
Speaker
They move at a certain pace and you get left out. So you try to say, okay, I focus on quality and all of this would have worked if there were no macroeconomic keywords. So it was a perfect storm of doing your hard work. We built our own nilliness, completely in-house. We built our own under-eating in-house.
00:15:41
Speaker
non-management system. He has the origination management collections, the underwriting, the scorecards. But then what we realized was, must be we ideally could have started with something that wasn't in-house, made those learnings, then built it externally. I think it's the start to perfect everything before we go to the market. Now it's completely changed how I think when I run a product, I just go to the market, have something that is friendable, see what the feedback is, keep on admitting powerfully.
00:16:06
Speaker
Now you have that every-tier process. Build a minimum viable product and then keep improving it. Yes, exactly. You're like, see what sticks, what doesn't work. We'll know then we'll keep fine-tuning. You'll need real-time market feedback because all your assumptions and assessments may be completely wrong.
00:16:22
Speaker
And you're asked goodness and astrologer if you don't have market feedback. And then it's up to your current followers to say you're right. Okay. Okay. Okay. Okay. Got it. Did you incur a loss at the end of it? Had you raised any funds? Like financially, how are you running it? Was it operationally breaking you?
00:16:44
Speaker
Today friends and family around Vinay is one more general round and there was a capital loss because there were certain acquisition offers which had we had the team agree to earlier we would have been acquired our capital would have returned to its IRR all of us would have been compensated but there were certain timeline issues and by the time we went to market there was this black swan event which kicked us and then eventually it became an equity hires, large employment offer and there I understand where the potential aquifers
00:17:12
Speaker
came from and why they fooled us. It was for us, ugly, unknown, and either the capital, safeguard their portfolio. But yeah, that was a personal loss as well, financially. But if you were to ask me, how would I have had those learnings without incurring those loss, and that would have been the ideal scenario.
00:17:30
Speaker
But the better scenario is we have all of those learnings. People tell me, this is how I'm going to make money by lending. And people simply assume lending is a feature. I think either they should be doing it with too much capital, which I think is still a far ask. They can throw around, see what sticks, and immediately build those learnings. Otherwise, lending is not as easy as people say. You can make money, definitely, but you have to put in those processes, put in those fail-safe scenarios, and then only build those capabilities.
00:17:58
Speaker
You're saying that lending is a long-term play. You need to invest a lot in good risk assessment systems for which you need to first generate data which can train your risk assessment systems. And the long-term value doesn't happen in the first loan. You need to start with small loans and then upgrade them to bigger loans. And therefore, it's not something which is just like a plug and play and start earning money immediately. It's not that kind of business.
00:18:23
Speaker
Typically, there are two types of spending models. One is your direct-to-market models. You don't have that much interaction with the customer. You don't know that much about him. And you're going to operate him like any other third party, any other user and evaluate him for credit worthiness. The second set of users are your captive audience. You've been working them and they've been on your financial, they've been on your, say, drug market place. So, quite some time, you have there and you have understandings that are close to their financial behavior.
00:18:49
Speaker
And then if you are in a wealth management or a financial management app, you have all of these financial footprints. So between an open ecosystem and a closed ecosystem, you always have better learnings in a closed ecosystem. You can ideally do, you can take a little more risks if you want to scale faster, because you have better learnings about it. So even despite all of that, see, imagine if one of your customers were being sticky with you for quite some time. Either they have defaulted. How do you handle that scenario?
00:19:20
Speaker
So they are your customer, they are defaulted, then is it going to be a collection issue? I want to lose a customer forever if I do not collect them the right way. If you don't collect, they're going to lose money. So end of the day, in a captive ecosystem, lending is when lending becomes a feature.
00:19:37
Speaker
Collections will be at the cost of customer relationship, existing customer life. It's a whole new factor for all of this. So when people say, okay, lending is how I'm going to make money, yes. Fair enough in India, you can make money and only in fintech and three-foot broad ecosystems. How you package it and sell it is what decides how you, how you want to stick with customers and skit. But the moment someone says, okay, my monetization stars, I've acquired customers, I'm engaging with them, but my monetization strategy is lending.
00:20:04
Speaker
Great, but it isn't as simple as it looks. It's a fine balance between collecting money, asking for repayments, because you can't have too much friction with your existing customers as well. And the moment a repayment is reported to a bureau, you're going to lose a customer forever. And this does not happen for just one user. It could happen across a portfolio of your customer. And even if you don't take it on your book, you're going to partner with another lending institution. End of the day, customers are only going to see you.
00:20:29
Speaker
Lending, yes, it's primarily a money-making machine here in India for analytics. Everyone has wanted to enter into Lending and there's absolutely no doubt and which is why most of the embedded financial and embedded lending players are scaling faster than ever. But it's a question of operations, a question of relationships, a question of underwriting constantly, iterating your scorecards and keeping in touch with the user. And we'll have to give where it is due for incumbent financial institutions, but they've been able to manage it pretty well to know.
00:20:55
Speaker
across multiple leaflets. So I think lending has a feature really broke, but it has its own set of problems that they need to address up front. So because of the value of data in building your risk models, that is the reason why a company like Cred is like maybe four or five billion dollars valuation because they have that data of credit card payment history of people on their platform, which allows them to make better risk decisions.
00:21:22
Speaker
No, I think the beauty of credit is they have put in a funnel where they're gatekeeping and sending in customers who have a score of 700 years and 50% above minimum threshold, which automatically ensures that you do not have previously fought history. And they are not first-time borrowers. The moment they have a credit score at Footprint, it means you're being responsible about their credit history. So that is the first part. The second part of the funnel, like you said, they are access to SMS data. They know where your money has been moving, where your expenses are. So they have financial user persona about you.
00:21:52
Speaker
All of this put together is pretty powerful to say I have been marketing this day for last six months. This is a credit score and every time you use a credit points top of your credit score, it probably also goes to credit right and they can potentially partner with someone to provide. It could be interest free, it could be even zero interest but for very minimal guess because even at that size of user base, they have really high quality users on paper
00:22:17
Speaker
with some tractor product free payments and they can potentially start giving lower tickets and I think that I'm a huge disbursement to under it right now and very acceptable entry levels which they reported.
Account Aggregator Framework and Data Management
00:22:28
Speaker
Right got it okay. So it's not just data the way they put in the customer acquisition funnel also ensures that there is a level of gatekeeping for credit worth views.
00:22:39
Speaker
So if you want to say, would it be the same for any other V2C facing up? No, I think Rindik has its own strengths and parameters basis, which they can find out better quality for us. Got it. Okay. Tell me how you bounced back. So you had your startup shut down due to the lockdown and the pandemic hitting and like multiple Black Swan events. Then what was your bounce back journey from there?
00:23:02
Speaker
So what we realized was we had the level of learning, because we were pretty placarded earlier, when we met a lot of other fintech founders, we are pretty surprised to learn that we put in a lot of effort and we built a lot of capabilities that people are building only then and they founded that.
00:23:19
Speaker
Those learnings could be easily transposed to either an existing startup where we would be able to add value immediately or solve those problems ourselves. So we had multiple offers to either join and build really good lending books for large consumers and companies. But because the problem statements were so close to us, we thought, okay, why don't we start focusing on it? See if we can build something around those.
00:23:42
Speaker
And the core fundamental idea was the moment your financial data is made available to any entity that wants it in a consent-driven format. You had a lot of these open finance players building capabilities focused on wealth. They would build verticals only focused on onboarding, only focused on wealth, only focused on credit. And that was the size of the market. We said data has always been a problem. Now,
00:24:06
Speaker
How can you provide a plug-and-flip tool to sit on users' financial data? What were the legal requirements? What are the complaints? How was the infrastructure on the ground? And that put us on to a path to start understanding how the entire Indian iPhone integration framework was. And we eventually became a part of Samadhi as a technology service provider and we started building on top of it. If I were to say those lending journeys, those
00:24:30
Speaker
learnings that we had an optical rate when you flip it inside a poetic sense, it becomes an open finance play that you provide the same learnings, but building on top of data and provide it to any other VDC facing entity. So give me a history of this account aggregator framework. What is it like for an outsider who has no idea what is this account aggregator framework?
00:24:52
Speaker
So, you are a user, you have multiple accounts in multiple financial institutions. Could be a banking current on their savings account. And if you look at it globally, there have been players like Flair, there have been players like TrueLayer, SoilDense, Tink, Affiliate, across the world, with EU or US who have done it. The idea is, can I allow the data that's residing in these silos amongst these financial custodians
00:25:19
Speaker
Can I allow the user to have the capability to share it to any entity that is requesting further in a consent to run for it? So just like how it's where UPI works, right? You pay, you click on GP. GP asks you, can I make this payment? You make a payment after getting your OTPs. In a similar manner, can
00:25:38
Speaker
In SIU, there are two parties to the Indian iPhone Diagramation Framework. Indian iPhone Diagramation Framework works a little differently from Global iPhone Diagramation Framework. It's here. It's governed. It's much more regulated. RBI has put in its regulations along with Ribbit, its IDM. Samarthi is part of it. iSpirit is part of it. So there's a lot of collaboration and market making that's been done with players here. In the US, until now, it was a market to run flatables, their own capabilities, and the broad best practices themselves here. RBI has done a whole host of heavy lifting.
00:26:08
Speaker
So it's like the same story as in the US UN Visa Mastercard, which are private companies in India, which is like a collaborative government private initiative. So the same thing is happening in the account aggregator.
00:26:21
Speaker
Yes, the only difference is, imagine if Rope was a framework. Entities could build on Rope. So instead of one player, RBA said, let me bring out a framework, let the entities are going to build those pipes, the infrastructure to move data from A to B. Let those entities be licensed by RBA. And those entities are called account aggregators, and there are subclasses of NBFC licenses. What does this say with the AI spirit?
00:26:45
Speaker
So there are voluntary bodies which put together the working framework, the day-to-day issues of how the entire account aggregation framework has to evolve. The ecosystem has to evolve because it's not just the pipes, right? There are two entities, major entities outside of just the pipes. By pipes, I mean the account aggregators who build the basic infrastructure to move data from A to B on behalf of the user.
00:27:07
Speaker
Now, you have the financial information providers. So, the entities which push in your financial information, the user's financial information into a centralized repository, which these pipes then move onto the entity that is requesting. So, on the left-hand side, there are entities which push in data. On the right-hand side are entities which request the data. And in between are the entities which sit on top of the AI framework.
00:27:29
Speaker
Now as account aggregation is issued by RBA, the entities that are going to pull in data necessarily have to push in data. So there is a requirement for you to share data for you to take data and you have banks like
00:27:42
Speaker
Okay, which incentivizes participation. If I am an HDFC bank and I want to know if I should lend to a customer who has an account in Kotak Meindra bank, then I must participate in this account aggregator framework to be able to access his data of other banks.
00:28:00
Speaker
And yes, if you want to access it, which will allow other banks to access data of my customers. Exactly. There was this earlier thought process that banks were debating, are we sharing, making our customer data such richly available customer data that we fought over years to build a market share, making it available so easily. But now the thought processes
00:28:24
Speaker
If you do not build value-added capabilities, someone's going to do it later and then you'll be left alone. So there is a mind shift change which is allowed banks and financial institutions to push and start pushing in data and also start making use of it immediately. Atchis Bank is on the quantization frame of post-testing their live industry and HDFC, ICICI. Now we are hearing SBA is going to go live. They've been in the talks so quite sometimes. So imagine
00:28:51
Speaker
The level of volume that's going to be made available, just like your pay was for peer-to-peer payments or peer-to-peer payments, digitalized. Now you're going to have a framework, you're not going to have a sort of a data revolution of sorts in the synthetic ecosystem, where everyone's data is going to be kept in a secure, consent-driven framework. And it's no longer sharing of your net banking passwords, no longer sharing of your user ID passwords. Everything is linked to your mobile number, just like your page. You put on your mobile number, it discovers what the accounts are.
00:29:21
Speaker
every account allowed to give an ODP and manually confirm the consent. The consent states who's asking for your data, how long they're asking for it, how many times can they pull the data within to them, what is the end purpose of it. So there is no misuse of data, right? Now you see so many bank statements moving around, people keep fudging it, people keep misusing the documents that are given in trust. Now, nothing of that sort is going to be made available.
00:29:45
Speaker
And that is what account aggregation is all about. Data that was siloed, data that was closed, financial data. Now imagine the number of networks, the number of relationships that can be made available once you actually scan it. And that is exactly, we are going to open it up and make it available to any regulated entity. So right now, data can be pulled away entities that are regulated by RBA, SEBI, TFRB, and IFBA. That comes to a large amount of financial information to use our entities.
00:30:13
Speaker
So, these entities which can pull data are pulling for essentially lending. That is the use case or are there other use cases? See, financial transaction data is in some sort and indicator of a financial behavior. Whether you want to do it for financial management, whether you want to do it for KYC checks, account checks, income checks, or whether you want to do it for understanding your credit worth, initial equity or lending potential.
00:30:37
Speaker
While lending may be the first use case, obviously the majority of the pools into the early era of account aggregation will be credit driven. After a point in time, once the consent and the purpose expands, there are more types of consent coming in. It will not be lending first. It will have a large portion of use cases driven around say wealth, driven around financial management, outside of lending. Why would a wealth manager need to know your banking data, your banking history?
00:31:04
Speaker
So now, how does someone approach wealth? You say, I have excess cash. This is what I want. You speak with someone, you upload some of these, even if they're like robots online, they look at your needs, your requirements, your goals and state something. Now, imagine if I were to say, these are my phones, these are my three cash flows, these are my liquidity patterns, these are my expense patterns. Everything has hidden in your banking transactions.
00:31:27
Speaker
right from your subscriptions to Netflix, right from your income, your expenses. Now I go to RobotRiser platform and say, this is what I am. This is my financial persona. This is my DNA. And depending on the type of financial advisor for me, when that advisor could be a robot advisor, people get a richer idea about you. And then,
00:31:46
Speaker
suggest the right wealth products. In fact, if you look at the way forward for account aggregation, it's not just banking data, non-disture costs or non-disture current and savings account data that's going to come into this framework. The roadmap states your credit card data, your health data. We have 22 different types of asset classes that will tie your product, financial asset classes that will be made available around the framework. But the closest to reality right now is just banking data.
00:32:12
Speaker
At the point in time, all of this will be made available and every FIU, every entity that wants to understand users better will just make it so seamless to come into this ecosystem and start pulling in data. Okay, so similarly, like an insurer could look at your banking history and then tell you what is the right policy for you, what is the right premium around, whether you pay monthly premium or quarterly premium based on your cash flows, they could give you much more personalized recommendations so that the decision making process becomes easy.
00:32:42
Speaker
Yes, exactly. And imagine the number of financial accounts always outstrips the number of people who have been featured in the financial ecosystem. You have so many Jansan accounts and all of which, even with minimal usage, you can understand what that person is all about. Might not be a complete idea, but you have something about that person. And that allows entities, financial service providers to take a core on that. And earlier, that was not available.
00:33:09
Speaker
If you like to hear stories of founders, then we have tons of great stories from entrepreneurs who have built billion dollar businesses. Just search for the founder thesis podcast on any audio streaming app like Spotify, Ghana, Apple Podcasts and subscribe to the show.
00:33:30
Speaker
Let me understand now the stakeholders in this. Now, one stakeholder is the ICICI, HDFC, which have the data. And the other stakeholder is companies who are looking to offer consumers some sort of products, be it wealth, be it lending as the first use case, or be it insurance.
Ecosystem Stakeholders and TSPs
00:33:46
Speaker
Who else is there in the stakeholders? Who's the data stored that central repository? Is that a RBI run thing? Yeah, see this, let me rephrase to give you clarity. The net has to be certified. The entities that are pushing in data and entities that pull out data have to be certified by.
00:34:04
Speaker
Either the account aggregators or technology service providers to state that they are entities who can validly push in and pull out data respectively. Now, the stakeholders broadly, as I told you, apart from the account aggregator, account aggregators, you're seeing you, you're on money. These are entities who have been building the pipes, putting it together.
00:34:24
Speaker
the base infra. Then you have, they're building the entire pipe from a source of data to the person who's asking for data that whole pipe is being built by them. So the pipe is at one end, right? So you can onboard financial information providers onto the ecosystem. And after a point from them, they become financial information users. So the other end, it just depends on which is the FIU that is going to come in full data. So the data
00:34:51
Speaker
pipe is available. And that journey, those pipes are blind pipes. In the sense, they can't see what data is going through them. That is the beauty of the entire ecosystem that's been framed by RBA, some of the entire dive sprits. And that allows for complete data privacy. Now, you look at the other ecosystem. There are crucial players in the ecosystem called technology service providers.
00:35:14
Speaker
So what they do is, to visualize, I gave you the equation, right? There's the information provider on the left, the pipes in between, and information users on the right. Every time there is the information user's end user giving consent, consent goes to the entity in which the mobile number is linked, pulls in databases, the consent artifact, and pushes us to the entity that is asking for the consent.
00:35:33
Speaker
Okay. So the data is not stored anywhere. It is getting pushed each time there's a request. Got it. Yes. There is no storage, which means it only reaches the entity which has requested for the data and the consent to which the consent has been provided. There is no misuse that can be done.
00:35:50
Speaker
And how is the access revoked? As you said, this is like a limited time access. How does that work? There are two types. One is you prefix. Every consent comes with a timeline. You give an expiry for the consent. Either a year or six months or one year, the consent automatically expires and you need to renew it.
00:36:07
Speaker
The other way is you have all of these concerns are managed independent of the client's app in the account aggregators app. Yeah, one of these pipe builders have their own customer facing apps which state where you've given your consent, do a change in duty, how long and even go revoke those concepts.
00:36:25
Speaker
So the entire data ownership is now back with the end user. And that is the idea for upon aggregation. So back to the other point I was stating, the other important stakeholder in the upon aggregation ecosystem is technology service providers.
00:36:42
Speaker
So, outside of the pipes, right? There are multiple use cases. So, every entity might not be equipped to build use cases on top of those pipes. So, that is why players like us comment. We say, okay, the pipes are available. We'll build the use cases on top of it, make it available in plugin play to either an FIU or an F5P.
00:37:01
Speaker
Either question data or make use of users' data and understand or engage or whichever is the use case that you want. It would be credit, wealth, marketing, whatever the consent allows for. So it could be credit or wealth, collections, etc.
00:37:14
Speaker
And that is where the ecosystem is going to grow. Because as per law, the append aggregators should not have, should not cross-sell services to the client with which they're working. People have different arms to do all of that. And PSPs as a brand of players in the ecosystem are trying to build value-added services on top of those data pipes. Could be, as I told you,
00:37:37
Speaker
And doing an account check would be an income check to basically understand user behavior. Or can you build engagement players? So if you look at one of the things that SIBO is trying to do, SIBO also provides plug and play PFM as a service, personal financial management modules as a service. There are multiple pieces, could be like a financial calendar, a cash flow tool, round up saving. All of these are independent engagement modules that can sit on your transaction data and it can do the work of engagement, engaging with your end user.
00:38:05
Speaker
and Sego does it, provides it as a single platform. The idea here is for retail B2C entities, the least common denominator for engaging in a financial ecosystem, financial parlance with their end-user is a PFL. It works in a, say, can work in a consumer interest company, can work in a financial market place, can work in insurance or robot does it, it can work in an NBSC. That engagements.
00:38:28
Speaker
Capability is missing right now to a large portion of the financial ecosystem. Imagine the case of a credit bureau or imagine the case of a financial loan marketplace. You go there, click on a request, you leave, ask a request for a loan, you leave. But how do you engage? Data is available. We make sense of that data and we engage on your behalf.
00:38:45
Speaker
UITI will provide the same days RSD case. So there are multiple use cases that are that can be built on top of account aggregated data. Could be vertical, could be someone could be focused on building place for a vertical like L for travel subject to consent, consent artifacts, or could be completely horizontal to provide a plug and play tool or master to whichever sector. So I think there's going to be a lot of evolution that's going to happen in this ecosystem.
00:39:12
Speaker
And you're already having differentiated players starting to come. And it's a question of when this hits critical mass and not is. Okay. Let me recap my understanding. Now the account aggregator companies, these are companies which are running blind pipes and they don't get to see the data. And these are, this is like an infrastructure layer. They're like API providers and the information givers and seekers can choose to directly connect with the account aggregator companies.
00:39:41
Speaker
or they can choose to connect via a technology service provider who does some value addition on that data. So for pulling data, the SIUs necessarily need to have a commercial and technical arrangement with the AAsper logs. The SIUs cannot, I mean, have to be tied up with the AAs for them to pull data.
00:40:01
Speaker
But after pulling data, how do you? What do you do? That is where evaluated service is on top of data. And the other way is TSP connectivity for connectivity also exists, like you said. So eventually the engagement will be with the open diagram address. The DSPs will say, I'll be a single point of contact for you for everything. We'll work back to back with the pipes. Your engagements will be with the pipes. We'll be with the open diagram address. Your commit hits will be with them. But we'll just do a whole host of material lifting for you.
00:40:27
Speaker
Because if you look at it, building pipes is a separate business. You have hand billion dollar entities focused on just building pipes. Be it play and be it and you have a lot of these pipes coming up in Southeast Asia. Be it in UAE, you have TAPI, you have lean, you have true layer in EU, you have saltage in EU, you have bioply in EU, you have lead. So the value-added services are
00:40:50
Speaker
Essentially, the larger pie in the money is available for you to build either vertical age or horizontal. That's depending on each one's thesis. The pipes are a separate business already, which is why the idea is the pipes alone cannot solve for it. The TSPs alone cannot solve for it. The pipes have to do their business. They have to build better quality pipes. They have to build the best-in-class quality pipes, which they're doing. They're a great place in the ecosystem. The TSPs necessarily have to build these value-added slavers on top of it to provide plug-and-play capabilities to any. Both work in tandem.
00:41:19
Speaker
that's been the precedence on across the globe. And a technology service provider can like a client signs up a technology service provider and then the technology service provider has the ability to directly interface with the account aggregator to work on the data which is coming in. And therefore these technology service providers need to get some sort of regulatory approval because you are able to access the data coming from the account aggregator. Do these TSPs need regulatory approval or anybody can be a
00:41:46
Speaker
They have to be certified and audited by Samadhi. We have certain norms where we have to comply with, and you have to be a part of Samadhi. You have to work with these account aggregators, and the moment you work with these account aggregators, you have to get certified. And you'll have to get yourself audited. There are multiple entities that have been certified by Samadhi to do this audit, the most technical and legal.
00:42:08
Speaker
So as per law, the end user has to interface, the FIU has to work with the up-and-back aggregators. And if it's TSP, you'll have to ensure that mom does not get violated. So whether you work in the right manner, whether your user flows to get data consent, whether work in compliance with what's being managed by Revit and some of these.
00:42:29
Speaker
So all of these are the necessary part of when you can work with an entity, when you can work with an FIU. And these constantly get updated as well because some of these practices are very unique to India and some of these we've been working by looking at best practices abroad. And it's a combination of market-driven and regulated environments that help us with the best thing that's possible for all of these entities.
00:42:51
Speaker
Okay. Okay. Now tell me something where there are multiple account aggregators companies. So like Rupe, UPI, these are also in a way in the pipes and there's only one single operator. Why are there multiple account aggregators operators? See, it's very similar to, uh, UPI switches or card switches, essentially.
00:43:12
Speaker
The more the number of players, someone will always be able to add value. And look at the guys who are coming into the ecosystem. You had a very deep distribution working with players for a long time. You've recently heard phone thing, you recently heard credit belts are playing for an error. Since cash-free is in the news, apparently either a plane or a flight for an account aggregator license.
00:43:35
Speaker
There are multiple players in the ecosystem who are looking at building these pipes either for internal use or for external use. And the competition between these players is exactly going to be one like how you play, you're going to have better user facing capabilities, better uptime, better engagement and the quality of data and the speed at which all of this is going to happen is just going to be better if you have more competition.
00:43:56
Speaker
The idea we can see RBA is pretty up the curve and they're pretty nuanced when it comes to all of these ecosystems. They've always ensured that there is the right amount of competition, there's always the right amount of regulation, the right amount of freedom for all these players to work with. And for them, customer data is paramount. If there is anything that causes some sort of issue, some sort of problem with customer data, then you need first guys to pull those players up.
00:44:22
Speaker
And for all of this, it necessitates competition between the pipe builders and cams also makes sense. And a lot of good quality players were coming into the ecosystem to build those pipes. I don't think that this is a winner-takes-it-all market, primarily because they're after 500 times, everyone has worn both.
00:44:38
Speaker
start looking at providing value-added capabilities. Couldn't be focused on wealth. Couldn't be focused on something. Couldn't be it where I retire without a certain partnership. I want to ask something here. So say, for example, like SBI is an investor in cash free. So let's say SBI decides to go with cash free as the account aggregator partner. Some other banks say SBI, yes, bank is going through Fort Bay. Will this data be shared between phone pay and cash free? Will a yes bank be able to access an SBI account holders data? Will these account aggregators
00:45:08
Speaker
to each other. Account aggregation pipes are independent in bus feature. They are just blind pipes. In fact, you'll never know what customer data is going through. So for an account aggregator to save, this user's data is also available here now. So every time a new SIU comes through, for example, if SBI is coming, is linking a user's account via account aggregator one slash account aggregator two, they will have to necessarily
00:45:31
Speaker
ask the users to link. Interoperability is still in works, it's not live yet. So right now, a pipe is blind to one data, goes through it and consents are to be done every time confirmed by the user and FIOs every time they come on hold. Also, if you're asking from the financial information provider,
00:45:49
Speaker
position, data that is pushed into the account aggregator framework, the same, there is no symmetry in it. So the same data that is made pushes in the airspank filter, sorry, into account aggregator one will be the same data that they push into account aggregator four. So there might be silly issues, there might be pricing issues, there might be uptime issues, but there'll never be data symmetry done that will be in that ecosystem. So there'll never be an account aggregator that technically has more data than an account aggregator. That's not the idea.
00:46:18
Speaker
There might be a few days about implementation issues, might be uptime issues, but never theoretically on data availability. So you would then actually need to work with all of them. And SBI would need to do a phone pay account aggregators and cash-free account aggregate collaboration. Not necessarily, which is why the better quality pipes. And SBI could eventually say, okay, I work with only one iPhone aggregator because all of my, all of this iPhone aggregator has wonderful uptime and they are better quality pipes. And it just makes sense for me to work.
00:46:48
Speaker
It's because there is no differentiation between A1 and A2. Logically speaking, they might just choose the better part. But what a model that are coming for TSPs like us, what it was, we do a multi-A aggregation play for an SBA. Instead of SBA working with multiple layers independently to see what works, we'll say we'll build it for you, just plug in there. So that is the basic connectivity module. So models like these will start in one way.
00:47:11
Speaker
Okay. Okay. Which is why then TSPs or technology service providers are important. They will make this plug and play for a client. As a client, I don't need to, as an information provider or an information seeker, I don't need to think of, okay, which pipe should I use to access this customer's data? That the TSP can figure that out. Okay. This customer's data shouldn't be accessed.
00:47:32
Speaker
Yeah, at least I'm just going to go between a conduit between you and after all of the legal documentation sorted out. And also, if you look at how typically an ecosystem like this evolves, the acceleration is done always by the players in the middle. So someone is wanting to build any added use cases, any added services that are extremely useful, extremely unique. Then.
00:47:54
Speaker
Consequently, usage on the account aggregation pipes will increase. Someone who's going to provide just the pipes and if they're not going to provide the capabilities on top of those pipes for specific procedures could be lending, could be engagement, could be... The moment you build all of that, the pipes naturally start growing faster than the ideal you would have. Which is why the TSP layer is super infected. And the TSP layer is where the money also lies and after a point in time, which is why you will see a lot of these AAs also talking about value added services either by bundling or partnering or building their own after a point in time.
00:48:24
Speaker
So, what is it that a TSP can do which an information seeker cannot do in-house? Why would an information seeker choose to work with a TSP versus doing everything in-house? As simple as someone who wants to make sense of that data, but in the case of an NBFC, so they want to take an account aggregated data.
00:48:44
Speaker
they want to break down the data and understand what is available and whether the client is credit worthy, whether the client is not credit worthy. Imagine if that NBFC is not a technical player or imagine even if that NBFC has minimal technical technology capability. Why would you want to focus on something that is not your core capability? Your core capability is lending, core capability is understanding user behavior. Now, all of those things are being made available by VSB. You see, look at the transaction string that hits your bank account.
00:49:13
Speaker
I'm just giving you one problem that needs to be solved which we go solving. Your transaction string is a garbled mess. Essentially, swiggy comes as swiggy, sometimes unbundled, sometimes a lot of transactions, sometimes there are multiple delimiters and different form of slasher hyphen. Merchant is upfront in one bank, later in a different bank string.
00:49:31
Speaker
All of this says something about you. Whether it is an investment, whether it is an expense, whether it is a subscription, who makes sense of it? Why should an NVFC focusing on lending? Why should an advisor, robotizer, wealth focusing on lending? Why should they even try to build these capabilities?
00:49:47
Speaker
a transaction categorization which is an extremely important part of open banking on top of account aggregation, making sense of that data. And once you have sense of that data, you map behavior and you constantly say this part of your portfolio is doing this. Imagine we do it across different entities, different clients. Yes, there will be no in exchange of data within one and the other, but this learning should be available for us to understand this user in this bucket is doing it.
00:50:14
Speaker
So are those learnings, can we even transpose those learnings to another client? And what we essentially say is we do all the heavy lifting right from complaints, the data availability, the cleaning up of data, behavior mapping, provide all of this as a single dashboard. Why is this even your business? You take care of learning, we just provide intelligence and we'll provide all of these capabilities to you, you focus on what you're good at.
00:50:35
Speaker
Got it. Because the raw data would be like, say, maybe a customer's bank account information would come in the form of 10,000 rows and each row being one transaction, which then for a company to make sense of, they would need to set up a whole team of data engineers and machine learning.
00:50:52
Speaker
Yeah, you're getting access to data as the least of some client's worries. What do you make sense of it later? You clean it up, sanitize it, understand it, break it down. Then on top of it, you provide use case relevant insights. So it's a whole different journey for someone to do it. And in fact, as I told you, the guys closest to the business, the account aggregator pipes, similar players in the US are used in their use.
00:51:15
Speaker
They're essentially focused on only building the pipes when they're so close to building the value added services. They focus on building this because that becomes a different business altogether. Different businesses in both these areas are flourished, independent of each other. Got it. So you are able to provide something which makes it easy for the lender to do a credit decision.
00:51:36
Speaker
It could be for any beta-spacing work. See who does this. We are hot. Overall, the thesis is, finance is no longer a separate vertical. It is an embedded feature. Now, can we do something with users' data on top of connectivity? We provide, as I use, the capability to pull in data via our connectivity module after the data is available.
00:51:57
Speaker
Users can do, clients can do things with that data to either engage with their users or understand the users better. Which is what we call as an engagement experience layer and an insights layer. As I told you, experience is pretty much VFM as a service broken down into independent modules, personal financial management. If you look at a retail facing new bank like a Chime or a new bank, and it is basically apart from the core and from account opening, et cetera, it is
00:52:22
Speaker
multiple independent modules put together, calendar, roundup, saving a health score, monitor rules. All of these are experiences that can be modularized and made available to any other entity that is an FI. It could be an NBFC, it could be a robot, vice-versa. Why should it only be
00:52:37
Speaker
within a retail facing new bank, etc. So what we said was, we'll unify it, we'll provide this both APS and SDKs, and we'll make it plug and play on top of data. We'll do all of the cleanup, we'll do all of the standardization prior to transaction categorization, we help engaging with the users. As I told you, one of the biggest problems right now is how do we engage with the user in a financial environment, and we'll make it available. And there's constant feedback that's coming in. That is one part of the equation.
Faygo and Seego: Service Offerings and Market Positioning
00:53:02
Speaker
Now, can the same data, the same
00:53:04
Speaker
data that's broken down the yield of identifying user behavior. Can we provide insights to the client about their users for very specific use cases? As you rightly said, broad insights, collections, and recoveries. So when does the liquidity pattern break? When do these guys have lesser expenses? How can they target for collections? And more importantly, wealth-related, who are users who are net worth climbers, whether they have exposure to insurance, whether they have exposure to mutual funds,
00:53:30
Speaker
All of these sit within your marking reforms. All of this can be massaged and provided as actionable insight. And then of checks, one-time events, if you want to look hit up users back one night, if an ad is working, what is his or her salary is. All of this hits your background is made available. And same as the case with account check, if you want to know this account, it will answer someone's thoughts.
00:53:50
Speaker
Under the intelligence slash insights buckets, we have two use cases. One is event-based, six the other is insights to monitor portfolio and user over a field of time understand what his or her trends are. So, Figo provide this capability, this platform for connectivity engagement and understanding user behavior via insights as a service. Can you give me an example of a hypothetical business that would be using Figo and the experience of the customers of that business?
00:54:20
Speaker
So imagine you are, let's take the case of a small bank right now, it will fit their mobile banking applications. There is no engagement, right? You hardly can do much, no. Or is there a small NVHC? Both of these entities are fine. So is there a requirement to engage with the users better? After they get the plug-in door connectivity module where they can ask the users for consent and pull-in data, they can embed our
00:54:43
Speaker
TFM modules are experienced here. So on top of the data, users' independent data in their existing mobile banking app or in their existing app, they can provide for community-side interfaces like ground-up savings, health score monitor, basis of financial transfer, cash flow management tool, an instant CD-Fox, a peer comparison score, so where they reside compared to proxies with similar parameters. All of these are capabilities. Every time there's a transaction that hits the back foot,
00:55:12
Speaker
the experience layer engages with the end user on behalf of the bank or the end user. So it's the problem statement for the client is I want to first get someone's data and I want to engage with them better. This could be the right solution for them.
00:55:25
Speaker
If the problem statement is, I want to get data and I want to understand better about them and better about my portfolio of users for credit-related insights, then you do the same thing straight to the purpose that it is going to be for loan credit. Pull the data and on the data that is folding in, we break it down. And for example, you're at digital lending, but if you roll it across 5,000 users, you identify 2,000 of them are in between jobs. After that, half of them are white-colored or blue-colored basis to stimulate your credit settings.
00:55:55
Speaker
your bank accounts, then all of them will identify if someone has a credit cushion or cash cushion within their bank. When there is cash residing, they are just good for a shorter tenure. So you can give a top up loan and there will not be a problem. And you can also identify with some not good for loan customers. And after good for loan customers, you can also see if there are potential competitors sitting in your wallet as early as the previous spring that it's the fund aggregator full-in data. So all of these insights
00:56:23
Speaker
our end of the day sit on top of financial data. And that is where we provide this capability. So it is very use case specific and use case driven. Okay. Okay. Okay. For you, these are the two big, biggest use cases that your customers use. So personal finance management and ring insights.
00:56:38
Speaker
We have had players working with wealth-related insights. Some of them focus on collections. So we brought here three types of insights. One is the risk and risk of credit worthiness. You have collections and fraud. And you have wealth as a separate capability. And you also have checks, income checks, and account checks. Now, what are you trying to do? What is the account check? Income check, I understand, which would be like estimating what is this person's income.
00:57:05
Speaker
And also employment, what does an employment rate does? How long have you been in that company? How effective it gets on your backbone.
00:57:10
Speaker
Now, if you want to talk about account 6, does this bank belong to you? Instead of doing a penny drop, you can just pretty much full and identify this bank as yours. How long is it back fills in dollar with all of that will be available. Now, the other thing to see was while we sit on top of account aggregation, the same capability pretty much works on any transaction string. Could be a forward banking string. It would sit on top of a bank's own data. So we made it agnostic while in focus on account aggregation, we also made the
00:57:40
Speaker
provided the option to entities who have their own access to data to plug into it. It need not even be an FIU, it can be a bank, it need not be an FIU that is working with our connectivity, it could be any kind of utility market. So the personal finance management thing, is it a monetizable business for the client or is it just an engagement
00:58:02
Speaker
It is actually the next step. See, from embedded exteriors, the journey to embedded finance is actually pretty close. The moment you understand your user better and engage with them better, you can sell the product. Let's take the case of Roundup saving now. So you are in a, say, a consumer tech company or constantly been spending
00:58:20
Speaker
You want to provide round-up savings after a point in time. Those virtual round-up savings could be provided as a discount or a coupon. Every time this positive financial behavior that the user does or every sprint that happens, you can reinforce by providing a coupon. After a point in time, imagine the round-up savings that is virtual.
00:58:37
Speaker
can, after a point in time, become real. You can push a digital asset or a gold, a digital gold, or a mutual friend. Any security wire wound up saying, no, this product or service takes different form across different services, different financial experiences.
00:58:52
Speaker
Round-up saving might be different versus financial. An example of round-up saving could be, for example, say on Swiggy, you place an order for 375 rupees and Swiggy tells you that the additional 25 rupees, why don't you...
00:59:07
Speaker
save it in some instrument. And that instrument, when you invest in that instrument, could be an SIP or whatever, that Sugi earned some pennies. Imagine if you're doing it in an ecosystem where you're not constantly spending. Imagine this can be done with an NVHC's app. So every time there's a spend happening, if an NVHC wants to engage you, the transaction data consent has been given. So you can build those experiences, could be around every time. You don't actually see a financial lender or digital lender's app engage as a user that often.
00:59:37
Speaker
Now, imagine this is made available every time you save. Now, all of the savings are virtual because you provide it to the earth. It won't be the fee from right now. After a point in time, we can host a marketplace of providers with a push-up, say, ready to go out and work. It's less than after a point in time, right? So right now, we provide this. A user comes, experiences that the industry can either give a cook-on or a discount code or whatever and a snatch on that processing fees. After a point in time, they can push a product right. It just, engagement is the beginning for
01:00:07
Speaker
We need a financial player to work with the user and be relevant to the user's lifecycle, transaction lifecycle. After a point in 10, once you understand and engage, then you can push a product. Okay. Tell me about the journey so far of buildings ego. How did you figure out that this is the space you want to build in? How did you...
01:00:28
Speaker
get yourself that regulatory approval. We are not an account aggregator. First of all, what we realized was the pipes after a point and then we'll get from audited. And the value added services is where your engagement, your money and the gig can build all of those capabilities on top of it.
01:00:44
Speaker
And in fact, the first problem that after it upgraded, we had to unfortunately move on. The first problem that we were trying to look at was what were users, what were borrowers doing for this specimen. If you look at how typical lenders identify what they're doing, either it would be via users SMS data or if the app is there or...
01:01:03
Speaker
If the app is not there, they try to provide a top-of-flow and ask them to re-afflate for which they may get backing data and see if they can give a top-of-the- ground to get. But after that, find it's a blind spot. You don't know what's happening. Are there any red flags as he's swiping the money he's not out? We said, is that a framework where we can monitor the user constantly and provide this capability to lenders?
01:01:24
Speaker
That was the starting point that led us, that search led us on the contact indication framework of global players. We initially thought, why not build this for players for an ecosystem outside India? Because the pipes were pretty much available. But what we sort of realized was, see, when you're building a capability, when you're pretty much making a market here, you always have to build a boat before the sweat comes. There's no point in trying. It might not be the right boat. It might be wrong. You would not have received as much rain for as much water as you expected.
01:01:52
Speaker
pretty much had to be verbatim. This allowed us the timeline to do an experimental lot of it. So we started little early. We did a lot of experiments in our form of transaction data, built our own categorizations in the earlier partners with different players. We said, what is the right thing? Should we look at insights first? Should we look at experience first? Then we started looking at insights, then we started
01:02:12
Speaker
hearing our clients during POCs saying, you know, we want experiences. This is what, in fact, when we went to onboard earlier, I mean, POCs, we realized people were more interested in experience. The first set of audience. Then we said, okay, let's talk about embedding experience. Now there's a lot of interest from lenders for insight. So actually what is evolving is there a certain set of clients who
01:02:35
Speaker
look at engaging with the users better, guys who have solved for either acquisition or for monetization, but they don't have written general engagement. But for other players, it's a little bit different. So use case sort of differs. And then we initially worked with a couple of account aggregators to understand what it was, then became a member of some of the parts created in their workshops to understand. But the learning curve has been pretty long. In fact, we've been part of the ecosystem when we've seen multiple Russian changes as well.
01:03:03
Speaker
And all of these changes, we've been accommodating the priority field. And right now, I'm a part of some of these technology service providers' steering committee, where we are in what's to the ecosystem. We constantly keep updating what problems we face. Because what is pretty clear is that this ecosystem will grow only if all the players in the ecosystem, only if one stakeholder in the ecosystem, collaborate and work.
01:03:25
Speaker
This is not a winner takes it on market because the use cases are multiple. Okay. Did you have money to fund it? Because you also wiped out your capital when after credit got shut down. I think what people realized was after credit, in fact, for quite a few months, the founders did not take salary. They said, we run it, we run the show and see what happened. But you can't fight against fool with nuts. And we were pretty clear that we did not want to build a product before we did the wrong work.
01:03:52
Speaker
But we had to go quick to the market. But upon the aggregation and financial infra, it is typically a slower ecosystem compared to any other SaaS player or V2V or V2C players. Because there are so many moving parts, complaints, parts legal parts, which you have to take care of. So we started initially working with a free seed funder, so on and so on. He said, let me just take the capital, HMI family office. That was a year back. This was more than a year back.
01:04:16
Speaker
Then when we built out all of the Saturday building this hour, there were inborns for our seed. In fact, there are a few acquisition offers early on. And some of those investors had part of our cap table as well. So Syson came in 3.4 lit the round. We have a pretty fin-dex heavy cap table if you look at it. We have 3.4, we have Syson. We have a few more micro-reaches like Firstx, Exynix, and special inverse. And if you look at our industry leader slash angel investor, we have obviously Kunal from Kale.
01:04:46
Speaker
We have Lalit from Groves. We have Swales from Bharat's Bay. We have both the F2P founders, Prabhu and the Madhu All Chinga still there. And what we realize is all of these players typically take a bit. All of these common guys coming together is a good validation, good signaling. That does not mean the journey is done, but it's a good starting point. And what has been the monetization journey so far? Have you started monetizing?
01:05:09
Speaker
No, we're working at EOCS current hybrid path pricing model is pretty simple. So we work for insights it's check, for income checker on check, it's just a transaction based volume based check. For insights it's per active user but subject for number of transactions. So you can't say it's a user that they are monitoring it. But once everything is 100 users under transaction, accumulate a number of transactions, we price it as one, beyond that it becomes an accelerated pricing.
01:05:36
Speaker
Help me understand what could be the revenue potential for Fingo. You know, what like some like that back of napkin calculations that if we work with one large company, this is the number of users. This is what the monetization is. So three years in, we can be looking at this revenue. And I mean, you know, when you would have been pitching to VCs, they would have also wanted to know what is the revenue potential. I want to hear that conversation.
01:06:00
Speaker
Yeah, so we'll first have to take a step back at how the entire ecosystem is made. Sigo Play completely depends on the developer integrating Sigo as a capability into their existing system and existing products. And once the book is done, every other engagement is only going to be incremental. So you essentially are working with a combination of transaction revenue
01:06:23
Speaker
By transaction event based, an account check or an income one time checks office. There is a lender that is trying to look at on board 100,000 applications a month. You don't need to first of all pull your or 50 rupees or 30 rupees or 40 rupees or look at their bank statement via PDFs. You just do an income check and account check each but first feature parameter.
01:06:40
Speaker
So, it can be a part of onboarding, it can be a part of positioning, it can be a part of post-loan monitoring creative scenario. So, if you look at the volumes that after pointed time that we're looking at, it could be anywhere between hundreds of clients, would we, each user, each client, we expect some of the largest clients that are doing QC's with half millions of end users.
01:07:00
Speaker
And right now we're going live with a percentage of it in some time. And every client that we're working with, obviously on daemon, they're not going to be rolling it out across. Comfortably, by the end of this year, we're pretty sure we should be having our own time when the rendezvous is either monitored or insights provided about in the end by the end of this calendar.
01:07:17
Speaker
And how much should we be earning per these 5 million users, like an average number. It depends on the service, whether it's an engagement service or connectivity or insights. So insights are pretty non-linear because you are just more of a transaction. Anyone can fool in.
01:07:32
Speaker
I think, comfort to me, by the end of three years, we should be around 30 minutes. And who are the competitors in this space, companies which are doing similar stuff as you? Everyone's taken a position on top of data, whether it is focused only on, say, SME focused, credit, gems, right? So there are players who are focused only on the related capabilities on top of it. In fact, some of the players who are PSPs themselves would be low in management, or low in organization.
01:07:59
Speaker
They can embed Figo as a capability. But broadly speaking, there are a handful of players who are trying to do either on-talk of data, insights focused on credit, or some of them building free-built journeys. I am here to see someone building the PFM as a service on top of natural data. Basically, there is so much you can do with the data, so many different ways to cut and dice it that each player can
01:08:22
Speaker
Project completely vertical is for an industry or a new or a sector or so i think that's the value added so it's a split that's one to one after one that much faster. And that brings us to the end of this amazing conversation.
01:08:36
Speaker
At this point of time, I'd like to make a request. I want to know what you think about the show and how we can improve it. Do you have suggestions? Do you want to discuss your startup ideas? Is there any way in which we can add more value to you as a disster? I love reading your emails and suggestions. Please write to me at ad at the podium.in. That's ad at the t-h-e podium p-o-d-i-u-m.in.