Cracking the Code of Financial Inclusion: Insights from Nandan Nilekani’s Digital Playbook

NOTE: This is a transcript of the interview including questions by the host and responses by the interviewee. Human eyes have gone through the script but there might still be errors in some of the text, so please refer to the audio in case you need to clarify any part. If you want to get in touch regards any feedback, you can drop us a message on [email protected].

Okay, thank you very much. So, about 14 years ago, in between my previous journalism job and my current stint as an entrepreneur in the media space, I was hit by a lightning strike, because of which I reached out to Nandan Nilikani and I said I want to come and volunteer with the Unique Identification Authority of India, which was also known as Aadhaar, which I did for about eight or nine months. I think a lightning strike hit him also and then he agreed to accept me as of the volunteering task force.

So, it gives me some insider knowledge and insight into how this whole ecosystem is developed. So, I thought I'll start there, Nandan. We've raced ahead, we've talked a lot about, you know, the account aggregator system, the ecosystem of digital financial inclusion, but tell us about the building blocks, maybe, you know, digital identity, digital public infrastructure, going on to digital financial inclusion and so on.

And tell us about the building blocks and how those building blocks came together and how they matter, even as we look at this whole ecosystem today and where perhaps each of us can find a place within that, whether it could be entrepreneurs or regulators and so on. And let's use that as a starting point.

Sure. So, I think it began in 2009 and I got this offer to join the government and I had a one page, half a page, one paragraph saying, give every Indian a unique ID. How and all, nobody answered that question.

So, we came up with this whole biometric deduplication and so on. And by the time I stepped down, we had reached 600 million and now we're at 1.3 billion IDs. And the important thing was that A, it was designed to be an online ID, which in 2009, like was a bold decision to say everything will be online.

This is long before all the stuff. And we allowed online authentication, online KYC. Online KYC in turn allowed you to open bank accounts and get mobile connections.

That KYC, then once the bank accounts were opened, you had the Aadhaar payment bridge, you could do direct transfer and then people started, you know, doing direct credit. Today, India does about 300 billion dollars of credit into bank accounts. And it's a hot topic in elections where everybody offers a scheme which transfers money into your account.

So, I think all that happened. And then, of course, post that, NPCI took up the challenge of building a new payment platform. They built UPI.

So, I think over the last 15 years, we have built layer and layer of this. The important principle was population scale, small transactions, frugal engineering, making sure that you had public rails and private innovation. So, from right from day one, the philosophy has been, we lay those minimalistic rails and then allow private innovation to flourish.

And today, the estimated value of companies created on top of DPI is about 100 billion dollars in market cap. So, I think that was the whole philosophy there.

So, tell us about at what point did the concept of a unique identity start merging with financial inclusion or delivering a financial service?

I think from the beginning, I remember the first phone call I got.

And why as well?

Yeah. It happened because the first phone call I got when this job was announced was actually from the RBI. And Ms. Usha Thorat was the deputy governor and she called me and said, this is exactly what we need. Because we need to have an ID to open a bank account. And they were trying to get people to open bank accounts. But because of KYC norms, they couldn't open because the people didn't have an ID.

So, right from day one, you know, RBI was a huge enthusiast about this. And that's how the EKYC came about. EKYC led to bank account opening.

And India has opened about 700 million bank accounts since the last 10 years, especially with the Prime Minister's Jan Dhan Yojana programme. So, once you open the bank account, then you had a way to transfer money. So, we use Aadhaar to transfer money.

Once you had a way to transfer money, you had to withdraw the money. So, Aadhaar enabled payment system and I will withdraw money. So, all that was like layered on.

So, right from day one, the financial inclusion was part of the overall architecture of this.

And I'm going to fast forward in a moment. But how does account aggregator then come in? I mean, it's another layer. But where did this layer come up?

Well, account aggregator came because once we had everybody has a bank account, everybody has an Aadhaar, everybody has a mobile connection, everybody can pay with UPI at high volume, right? 18 billion transactions a month, including Google and phone pay and all that. Then, is there a way for people to take advantage of the digital footprint and monetise it for their own benefit?

Because the way data has evolved in the digital world is it's aggregated by big guys who then use it to show you ads or whatever. But if you could flip the data around and put it in the hands of individuals, and they could use their own data to advance their lives, then that's another possible revolution. So, that's how the account aggregator idea came up.

But in all fairness, the original account aggregator idea was RBI. And RBI came it out, what some of us did was we wanted RBI to make it a digital first platform.

Okay. So, let's talk about 2016. Why was that a turning point?

2016 is the defining year in the history of India's digital financial transformation. And I'll tell you why. April 4th, 2016, India reached 1 billion Aadhaars.

So, we reached a landmark on ID. April 11, 2016, UPI was launched. So, that laid the foundation for that.

September 5th, 2016, Mr. Ambani launched Jio. That laid the foundation for massive mobile penetration. November 8, 2016, currency was removed, so-called demonetisation.

And December 29, 2016, the Bheem app was launched. So, all these things happened in the space of nine months. But one more thing happened in 2016.

In September of 2016, the RBI issued the master directive on account aggregator. So, we're all here starting with that directive. So, you know, it's a phenomenal year.

So, if you were to now fast forward and look at what is being served and what is not, I'll come to what is not as well. And if you look at the ecosystem of, let's say, we talked about small and medium enterprises, we've talked about individual loan borrowers, maybe large loan borrowers. What's the segment in your mind that's been charged, served well? And what are the segments that have not been served as efficiently or effectively?

I think our goal is massive democratisation of finance, right? I mean, just like UP and all made payments democratised to a billion people, is there a way to get access to financial service for a billion people? And that's what we have to do.

Because historically in India, credit went to big guys, because it was easier to make the assessment of risk of a big guy. But small guys didn't have any, there was no information, right? So, you had massive knowledge of symmetry, which meant you couldn't give loans to small guys because the transaction costs are very high.

So, what we're doing is removing transaction costs and eliminating knowledge of symmetry. So, a small guy, a vegetable vendor or, you know, a small business can get access to credit. And that is enabled because of the AI infrastructure.

So, the democratisation of credit, where every individual and every small business can get access, is a huge level. It gives the opportunity for everyone to participate.

And what would you say are the lessons from the successes in as much as they can be applied to future innovation?

No, I think there is a playbook which we have evolved over the last 15 years, which is create it at population scale, make it frugal and low cost, make it have small transactions, sachet transactions, open it up with APIs, create infrastructure for private market operators to use your infrastructure on an equal opportunity basis, so it's not biassed, everybody gets access to the same thing, and then enable things to happen. And once we get all this equation right, then the magic happens.

The thing is, it takes many years to get there, right? You know, Aadhaar was conceived in 2009, reached a billion people only in 2016. UPI was conceived in 2013, launched in 2016, today we have, you know, 18 billion transactions.

Account aggregator was conceived in 2016, today we are at takeoff point. So, it takes many years for all the pieces to fall into place because there's government regulators, market operators, capital, technology, there's so much happening, right? So, it takes time to get all the ducks in a row, but once the ducks come in a row, then it takes off.

And I think what's happening here, and you can see it by the scale of this event, you know, 1600 people or whatever, you can see it by the quality of solutions that people are offering. This AI system has now reached takeoff, it's doing already 8000 crore loans a month, that's enormous, a billion dollars a month of credit. So, I think now it's at a point of takeoff.

Obviously, you have to fix things at scale, the banking systems have to work well, and and so on and so forth. There's a few things we have to get. And it's not just about banking, it's also about other areas.

The beauty of the AI is that it's actually for the financial sector. So, the way it works is, the decision on AI was taken at the level of the Financial Stability Development Council, FSDC, which has all the regulators on it, and RBI is the lead to implement that. So, that's why we're seeing now capital markets, gentlemen here from SEBI, we are seeing the use of that.

Now you can, you know, people are using it for personal finance, for that I had to get the statement of mutual funds, which I get from GAMS or KFIN, I had to get the details of my DMAT for shares, I get that from, you know, CSDL or NSDL. So, suddenly, a lot of capital market applications are coming. So, you're going to see more and more of that, insurance tomorrow, pensions.

So, I think the fact that we have this AI infrastructure across financial sector actually dramatically expands the envelope of possibilities.

So, I'll come back to AI in a moment. You know, there's a mention about scams. One of the things about accelerated UPI usage is obviously the bad guys also now use it to drain money from your account by conning you and so on. I'm sure this is not something that was thought of earlier and couldn't have been perhaps. But do you see this problem and how do we address it? Because this is also something that can turn people away from doing those lubricated transactions.

No, I think as Kunal was talking about it, I think how do we, just like we have done accelerated financial inclusion, how do we do accelerated financial literacy? Because we know that this has never happened. 400 million people, you know, consuming UPI.

That's seriously large, right? So, and many of them are first-timers to this world, first-timers to a digital world, first-timers to a smartphone world. So, essentially, we have taken decades of development and crunched it into two, three, five years.

So, there are bound to be possibilities. And you know, the fact is we are seeing the scams, not just people who are poor and illiterate, very rich people, very sophisticated people are being conned by some policeman who comes on some TV with a sign behind him and giving away crores of rupees. I've heard of CFO, fake CFO convincing CEO to give money.

So, this is like, so I think the scams also have become very sophisticated. So, we have to address it. They're not technical, really.

They're more like phishing or other attacks. So, we have to address the literacy at scale, which I think market operators and the RBI and all the regulators will have to do. So, I think both will have to go hand in hand, but there'll always be a lag.

I mean, the whole factory is doing scams now.

So, you talked about regulation and let me ask you a somewhat broad question first. Where do you think regulation needs to now focus on in the context of digital public infrastructure and digital financial inclusion more specifically?

No, I think it's by and large in a good place and Sharad was mentioning that the reason what we have done in India is, and Rahul Martin is also here, we have essentially done techno-legal development. Since regulators and technologists have worked hand in hand to develop this infrastructure, like when we did Aadhaar, we also drafted the Aadhaar Bill. The Bill went through a huge amount of privacy and all that, but finally, we got the Bill out.

So, the Bill was not drafted in isolation. The Bill and the technology were built hand in hand. The regulations were done, the rules were done hand in hand.

The privacy architecture was, everything was done. Similarly, the account aggregator was a master director from RBI, which we then adapted to a tech approach. So, it was like we brought tech plus AA together or UPI was done with RBI in consultation, KYC and AAPS.

So, a lot of the changes in SEBI, ASBA, which NPCI did, which has changed the whole IPO business, was done with the SEBI. So, fundamentally, the beauty of India's story is that the regulators, the government and the DPI gang has actually enabled this to be done. And same thing will happen in the future also.

So, it's not like some crypto bros coming and doing some random stuff.

So, do you get a feeling that, when I look at some of the Reserve Bank's actions, for example, in the recent past, including our banks, non-bank finance companies, gold lending institutions, there's clearly a sense of excess and that excess is also enabled by technology. Is that something you think about?

Look, every society goes through these cycles where there's a huge amount of credit, credit becomes easier to get, then people borrow from multiple people, they do evergreening of loans. I mean, we've seen this cycle. And then obviously, the system wants to ramp down these things.

We just have to find this balance and make it happen. There's no easy fix on these things. So, I think we're doing the right thing by allowing markets to innovate and then reining in some of the more egregious innovations.

So, we talked about small and medium enterprises, and there's a big gap there. I think what Shalini said was 20% is addressed, but 65 million is a large number of total MSMEs. How do we, let's say, or how do people here pick up some of those data gaps and start working on them in order to enable that financial?

Nandan Nilekani

No, absolutely. I think you want to see a huge growth of democratised credit for small business. Because as I said, first, you need to address knowledge asymmetry.

How do I know the performance of a small business? And in a pre-AAA world, a pre-digital world, I have to physically go and inspect, and that's not viable. But today, if a small business can say, get my GST returns from the GST system, including the invoices that I ship to GST for credit, or go to income tax system and get my income tax returns, or go to my bank and get my bank statements, or go to my wherever, or I can say, or I can ask my auditor to give me a certified financial statement of my, because I have an accounting package.

Then look at my flow of UPI transactions to show what is my sales. Suddenly I can construct a 360 degree view of a small business from his data. And if that can be securely sent to one or more lenders, then the lenders can decide to give me a loan on the spot.

So amazing. So you got data, you removed knowledge asymmetry, and you made it real time. So that means all these folks who are relying on their friends and their family at exorbitant interest rates, now can access credit from the formal financial system at scale.

This is a revolution.

So let me ask you the mandatory AI question. So how are you seeing the application of AI in this space specifically? And then on to, we also wanted to talk about ONDC.

Sure. I think AI is getting a patch. I don't talk about it. I think we are seeing AI, obviously you have to use it to reduce fraud. You have to make sure that you serve out highly personalised solutions. But there's also another big use of AI, which is in language. Because India has this peculiar issue of language. So we have 22 official languages and several hundred languages. And how do we make sure we provide access?

So some of the work we've done in AI for Bharat at IIT Madras is to create open data for Indian languages, which can be used by everybody. For example, NPCI today is using these models to give voice actuated commands in Hindi or English for payments. So suddenly, if a person in India can speak in a language of their choice to a phone using a WhatsApp bot or whatever it is, and communicate either information, like the best agri information or tasks, send money, then suddenly you made it accessible to a billion people. So the AI also has a huge role in inclusion.

Are you seeing something beyond, let's say, LLMs and that kind of application at this point?

I always said LLMs are a commodity. You know, these four, five big guys are going to spend $200 billion building the next LLM. That's be my guest. I mean, you want to spend that money, that's your problem. But the reality is that it's all going to get commoditised. I think the real value will be the layer on top, building the applications and building the small language models using data specific to enterprise or to a government.

And that's where the action is. And I've always felt that India will be the use case capital of AI in the world. So you're going to see massive amount of AI based applications, and some of them are being built by people here.

I'm going to come to Finternet, which is something that you've been talking about. And I think this is all comes together quite well. Just a quick alert. I'm going to come up or open up for questions as well. So do keep questions ready and short ones if you have any. Tell us about the Finternet, Nandan. That's your latest project, so to speak.

Well, actually, the two guys who really did Finternet are right here. Pramod and Siddharth. I'm just the guy who talks about the empty suit.

The fundamental thing is the Finternet is a way to do the next level of global financial systems in a way that's universal, unified, and user-centric so that everybody can participate. And the idea is that use the tokenization technology, cryptography, which basic strength is that you can create a packet which is self-contained, which has all the information about a particular asset. And then that becomes transactable, which is what the crypto guys did.

But they tried to create money and all that. But forget all that. But if you can take, for example, today a bank cannot give a loan easily to a real estate.

But tomorrow, if we can tokenise that real estate asset, then you can give a loan to that. And then that loan can be sold between two players. So this is the idea to tokenise all kinds of assets, be it deposits, loans, money, or it could be real estate, art, horses, whatever you want.

So the idea is this will lead to a massive growth in financial activity. And this will all be real-time, high-value transactions, 24 by 7 across the world. So you can use the power of what the crypto guys had without the baggage of trying to create a parallel financial system because the internet is within the world of existing finance.

And it's fully regulator-friendly and so on.

And do you see, let's say, entrepreneurs being able to create... No, for example, DigiLocker seems to be somewhere part of that, isn't it?

Oh, yeah. DigiLocker was conceived in 2015. Pramod Ramsevak Sharma was IT secretary.

And they came up with that. And then it was built by Amit, who was a volunteer. So many people contributed.

So DigiLocker is an amazing success. You know, 350 million people use DigiLocker and keep every document. Today, when you go to the airport, all you get is your phone.

You show your ID or your DigiYatra or you make the payment. It's unbelievable what's happening here.

So what could be the tokenization, let's say, opportunities? And does AA fit in anywhere in that?

Yeah, sure. Because tomorrow, for example, you want to... One of the things in credit is invoice fidelity, right?

Because in many countries, just like people take loans from multiple guys, people give their invoices to multiple lenders and get loans. So then again, all the stuff starts. So you can actually tokenise invoices.

And then those invoices can become portable and they can be used for it. And you can make sure that because it's immutable, in the sense nobody else can change it, you can say that this invoice is now locked. Nobody else can use it again to give loans.

Stuff like that. So now I don't get all the use cases. But absolutely tokenization will play a huge role in the whole journey of digital transformation.

And it's a bottom of the pyramid. I mean, I know you work on the BIS, which...

BIS is a global land of papers jointly authored ostensibly by me and Augustine. His real mind is ostensible. So I think the idea is that the regulators are also globally saying, how do we go to the next level?

And there's a lot of patchwork going on. Somebody's doing some CBDC project here. Somebody's doing some blockchain there.

But is there a strategic architecture for this? That's what Finternet offers. And that's why there's a huge amount of global interest.

Right. Okay. Questions, anyone? Yes. Okay. I don't know if there are mics. But yeah, there are mics. But why don't you go ahead first? Just give your name, please.

Hi, Smita. How are you? I'm good.

Smita - Thank you. Clearly, what India has done in the last few years, as you just described, is unparalleled, even by world standards. If you see any other country, nobody has done it at this scale. And so if you see this vision of getting every Indian, the 1.4 billion people being economically empowered through digital and financial inclusion, and we are already on that path. What do you see as the challenges or risks on this journey going forward? What would you be worried about? What could go wrong?

You know, many things obviously can go wrong. There's the issue of privacy. There's the issue of security.

There's the issue of financial illiteracy leading to people getting scammed. There could be the digital divide, because people don't have access to devices. People don't have access to language.

So we have to systematically tackle all this. I mean, like digital divide. I don't see why we can't give everybody a phone.

If we're going to have a scheme to give everybody money, we can also give everybody a smartphone and be done with it. So there are many, many ways to fix that. But we have to stay on course, because we are sort of halfway on the journey of transformation at scale and deal with all the challenges as we see them.

Okay. There was someone here. No. Yeah, go ahead.

Sneha from Jio Financial Services

Hello. Yeah, am I audible?

Sanjay from SBI

Yeah, yeah, yeah. Go ahead, quickly, quickly.

Sneha from Jio Financial Services - Yeah, I'm Sneha from Jio Financial Services. So we spoke about UPI. We spoke about the DigiLocker.

We spoke about account aggregators. So what are your views on the ULI scheme and where will the AA fraternity and all the peers that we are seeing today, all the products that are coming up, are going to fit into the ULI infrastructure?

Well, I haven't really... don't know too much about what's happening on the ULI side. But obviously, we'll have to find a way to integrate all this together.

But I don't have specific information on that.

Sanjay from SBI - Okay, yeah. No, I'm Sanjay from SBI. Sorry. You wanted to probably share something about ONDC. I wanted to know what is the current status and what is the future like on ONDC?

It's a very...

ONDC is doing about 15 million transactions a month, 14 million transactions a month. And they're across categories. It's in commerce of products, food, and mobility.

So I think it should grow much faster. And we are going to be seeing more and more companies doubling down on ONDC. And the key thing is you need some big players also to drive usage, right?

Because it's really just a... it's a protocol at the end of the day. How do you create solutions?

So there are many people looking at hyperlocal commerce or quick commerce on top of ONDC. So I think over the next couple of years, you'll see a lot of that happening, which I think will lead to huge growth of volumes.

I think the Namma Metro...

Namma Yathri. Namma Yathri does today 140,000 transactions a day, all on ONDC protocols.

And Koshi, who runs ONDC, is also from the extended Aadhaar ecosystem.

Yeah, Koshi used to work at E&Y. He worked on Aadhaar. Actually, Koshi, I go the long way. He worked at NSDL 30 years back. So we go a long way.

And I remember, I think he could manage 100 million IDs on NSDL. And therefore, the next challenge was how do we build a billion ID database?

I think if you look at the story of financial transformation, what India did in capital markets in the 90s was amazing, right? We created a new institution called NSE. We created a new institution called NSDL. We dematerialised all shares. We created new registrars, you know, cams and all the KFN and all these guys. It's the real story that began with that. But it was more back office orientated. And it was, the volumes were not that big. There were only about 10-15 million people who traded in shares.

So then when you take that backend, which was built in the 90s, and then you add this layer, led to another revolution. I mean, if you look at what's happening today, we have many, many people, you know, broking, you know, discount brokerages. You have all these, Zerodha, Grow and Angel One and Dhan and all these guys.

So suddenly we have seen a massive increase on trading, all enabled by combining the backend built in the 90s with the front end built here. Similarly, in banks, the bank automation began in the 90s, right? Everybody, every bank went.

And that combined with DPI created the opportunity. So, you know, all this, if you look at the large story, you can see all the pieces all coming together.

Narsingh Rao

Yes. Nandan sir. Yes.

Go ahead. Yeah.

Narsingh Rao

So going with the today's…

Introduce yourself.

Narsingh Rao - So I'm Narsingh Rao and I'm a SEBI registered research analyst. I'm Narsingh Rao. I'm a SEBI registered research analyst.

So we talk about a 10 trillion dollar economy or a 20 trillion dollar economy competing with China. And that comes with the MSME or going with the today's theme. And it requires an inter-ministerial kind of data integration, whether it is through MSME registration or the UDME registration or the KVIC data, which has to be there.

And at the same time, ONDC has to be there. So going with beyond the Ministry of Finance, if it has to be integrated. So are there any sort of such sort of delegations, which will help the MSME to grow at a larger scale?

Well, you know, again, you know, the UDME registration for small businesses is, I think, in the ministry of MSME ministry and all this, right? There are many parts to it. And we have to get all these things to work together.

But I think in the meantime, see today, if we can get the AI system going at scale, if, you know, today we're doing a billion dollars a month, we can go up to 10x this in the next few months or years. Then, and we put more and more data on the grid. See, for example, the moment you put GST data on the AI grid, then suddenly 11 million businesses can use their GST returns and invoices.

You put the EPFO data on the grid, then 50, 100 million people can use their salary slip to get a consumer loan. So it's all about putting these huge things we have onto the data grid and opening it up for consumers and small businesses. So I would rather do that than try to work with five people to agree on something.

That takes more time. So I think over the next few years, we're going to keep putting more and more FIPs on the grid. And the more FIPs you put and you make them reliable, which is one of the challenges we have today.

If you make them reliable, then the FIPs will come. The use cases will come. And when the use cases come, startups and big banks and others will work and the cycle starts.

Okay, so we've... Okay, last question. We've run out of time, unfortunately. Go ahead.

Alok - Hello. This is Alok here from Dun & Bradstreet. So we see a lot of these stacks coming up in India.

How do you see this actually expanding outside of India that we actually become exporters of these services into Asian and other developed economies?

Well, actually, there's a lot happening there. We have a number of agencies we have created. First of all, our goal is to take the DPI philosophy to 50 countries in five years.

And Chanchal, who's here, worked with us on the G20 Task Force Report, where he was helping us and which I was chaired by Amitabh Kant and myself. And now we have a framework on how the world should think about DPI. So I think about 20 countries today implementing DPI in some form, not the whole stack that we have, but some piece of that.

We have a group called CoDevelop, which is a philanthropic initiative to bring capital to DPI so that you can fund projects. We have CDPI, Centre for Digital Public Infrastructure, which Pramod is a co-chair, which is at IIT, IEEE-ITB, which is helping these countries to implement DPI. And many, many, and there are many pieces there.

So we have created a global ecosystem of actors to take the DPI to 50 countries in five years.

Okay, so, you know, just to close, Nandan, I think, you know, when we talk about institutions coming together, and I mentioned Koshi for a specific reason, because you actually pointed out how he built capital market infrastructure or contributed to the building 25-30 years ago, and today is working on something very sort of transformational on the cutting edge. I think this is also a matter of people. I think the reasons all these blocks are coming together, these layers are getting added on so seamlessly because the people are willing to talk to each other and the people are willing to work with each other and respect each other. I think that's a very critical point, which perhaps doesn't get mentioned so much, but I can see it, at least from the outside.

Koshi is sitting right here, yeah.

Great. Thank you very much, Nandan. Thank you very much for everyone. And let me hand it back to the speaker.