India – Scaling the heights

Many startup entrepreneurs dream of their offering taking off like a rocket. For Abhinav Gupta and his co-founders of IndMoney, within months of the official launch in early 2019 they’d gone from zero to one million users within the first year of business, with seven million users after just four years. Abhinav explains the rapid uptake, “We came from banking, so it helped get traction because we were able to refer the app to our previous clients who we’d been handling. That gave us the initial group of users and set up a network effect where they found it really useful, and new in the market. Everyone was talking about it on social media and other platforms. Once we were confident that there was a proper product market fit and people were really engaging with it, we went on to do performance marketing, using Google, Facebook, and Instagram as a way of promoting and spreading awareness. That’s how it got such fast traction, because the product was very novel and we were able to scale it in a really nice way.”

Removing the pinch points

OK, so this is a man who has quite the success on his hands, with the INDMoney All in One Super Money App allowing access for Indians to invest in Indian Stocks, US Stocks, Direct Mutual Funds and Fixed Deposits – all easily and transparently. It’s what Abhinav now describes as a bouquet for the needs of IndMoney clients, who for the first time could enter previously inaccessible markets. There was a lot of buzz around being able to invest in US stocks such as Apple and Amazon in a smooth and friendly environment. IndMoney wasn’t the first to open the way for Indians to access investment opportunities, but prior to their offering, investment was mostly in the physical mode and there was usually friction and pinch-points in remittances, such as with the US wallet. It meant doing paperwork at a bank branch, or online registration that was clumsy and time-consuming. Digitizing the entire remittance flow with banks gave users a unique experience which they were familiar with through domestic transfers between accounts. Result? IndMoney is number one in US stock investment across India with a user base of 10 million and growing.

Getting a consolidated view

Abhinav started his career in India with HSBC as part of the private banking and wealth management division. It was traditional private banking and didn’t touch heavily on technology solutions. However, the problem statement for many clients was that it was difficult to get a consolidated view of all their investments and liabilities in one place. Clearly, when all holdings can be reviewed in the same place, then the client and their advisors can make solid data-based decisions. This was the nut that the IndMoney tech team set out to crack, and the result is the consolidation of 80-90% of all available data. Within five minutes of a client downloading the app and giving the appropriate consents, almost their entire portfolio will be in front of them: All the data from different mutual fund houses, alternate investment funds, stocks and bonds. It was what Abhinav describes as “a kind of wow feeling at the time.” With feedback from customers, the app was quickly fine-tuned and ‘intuitive’ UX-UI was important in this. “Download the app, press two or three buttons, give consent, and you have your entire portfolio in front of you and can start investing and transacting.”

Bridging data science and the real world

It may be unusual to find a top data scientist mingling with local artisans and market vendors, but this is where Saurav Ghosh found himself when being onboarded for his new job with Indian payments company BharatPe in 2024. While there’s some distance between AI, Machine Learning and street markets, the logic of getting to understand real world customer needs is inescapable. That’s because BharatPe has a vision of making financial inclusion a reality for Indian merchants, often at the lower end of the earnings scale. While I’m imagining ‘Mom and Pop’ stores and thriving small businesses, Saurav makes the correction that those tend to be at the top end of customers for the company. BharatPe’s user base is often more likely to be unbanked operators, or people who find it hard to get loans at decent rates and cannot finance their enterprise to expand and succeed further. Hence the valuable exercise of going out and seeing what happens when high tech meets real-world needs.

Interoperability is the key

So, what exactly does BharatPe do? The company serves a wide spectrum of merchants – ranging from large enterprises to medium and small businesses – through a single interoperable QR code that works across financial service providers. In India’s diverse payments ecosystem, which includes platforms such as Google Pay and Paytm, its service offering is accepted by all major payment operators nationwide. This is supported by QR codes, devices such as loudspeakers and POS terminals – sourced from third-party suppliers and provided to merchants at onboarding. Beyond payments, BharatPe also extends working capital solutions to help merchants establish and expand their businesses. As of October 2025, around 15+ million businesses were registered on the platform, with the company also reporting its first profit before tax.

People with very thin credit histories

At first glance BharatPe’s proposition appears straightforward. Why then appoint a Head of Data Science with a mandate to ‘establish thought leadership through top-tier industry events, including AI/ML roundtables with Databricks, Snowflake, and Cloudera, thereby enhancing brand visibility among industry peers’? And what role can Artificial Intelligence and Machine Learning possibly play in the context of a small merchant – say, a street vendor selling mangoes?

Saurav explains that for many BharatPe customers, seeking credit is either a first-time experience or follows a series of rejections from traditional lenders. Most have little or no formal credit history, and some have poor records. Local moneylenders often remain the only alternative but their loans are typically expensive and unreliable. This underscores the need for a more structured and data-driven approach to evaluating new borrowers and monitoring their progress over time.

“Many of these customers have very thin credit files, which makes banks and NBFCs reluctant to engage with them,” Saurav notes. BharatPe addresses this gap by leveraging transactional data to gain insights into a merchant’s business. Where bureau data is available, it’s combined with platform-level information to develop robust credit risk and underwriting models. These models form the backbone of BharatPe’s lending approach, enabling access to working capital for merchants who would otherwise remain excluded from formal credit channels.

Importantly, the impact extends beyond lending. By helping merchants secure credit through trusted and transparent channels, BharatPe supports their integration into the formal economy. In doing so, merchants not only receive financial assistance but also build a credit history that strengthens their prospects for future growth. Over time, this creates opportunities to scale their enterprises while contributing more visibly to India’s economic development.

Everyone loves gold

From mango sellers, to … well everyone. Because “In India everyone loves gold, right?” says Prashant Singh, Head of Product at Jar. But gold is a physical asset and as he explains you can’t buy a rupee’s-worth of physical gold, because that would be just a ‘shred’. “You need, let’s say 10,000 Indian rupees (roughly $110 US) to maybe buy an earring, and that 10,000 rupees is a big deal for most people in rural and tier two segments of the country. India is an amazing place. Honestly, having seen a bit of the world, I can tell you, this place is an anthropological laboratory!” Prashant expands on the cultural importance of gold in many people’s lives, characterizing it as something of an obsession, with Indians acquiring gold ‘by the truck load’, and people who don’t have gold dreaming of saving to buy it. This can often be for ‘splurging’ on weddings where it’s important to show prosperity through the jewelry worn by the bride.

The digital piggy bank

But how does a vast population of relatively low-earning people get access to a share of those glittering gold bars that play such an important part in the national consciousness? The answer is bit by bit, in the form of those tiny shreds that become possible through virtualizing the saving of the precious metal.

This then is the background to Jar, the Fintech started by two friends Misbah Ashraf and Nishchay Ag, from Delhi and Bangalore respectively. Having both been successful startup entrepreneurs, they were looking for new ventures that might have a special fit with Indian culture. Prashant was an old acquaintance of both and was around for a lot of the ‘noodling’ of ideas of where to go with a new business. The persistent thought was of a digital piggy bank.

A piggy bank for gold

In the years following college, Nishchay had observed that there were few financial products aimed at people like him who were earning ‘decent middle class money’. Banks tended to wait until earners were five to ten years into their income cycle before addressing the market with savings or investment products. It was only then that offers of wealth management, equity, credit cards and so on started being made. So in the meantime, what was the best and most common way of putting a little income aside for savings, a rainy day, or a marriage? The time-honoured savings jar approach called for investigation, and in particular the concept of a digital piggy bank. The genius part was in linking the piggy bank with the Indian love of gold. If everyone wants their share of wealth through gold ownership, then give it to them, digitally.

Ownership and custodianship

So, real physical gold, stored in a vault from which anyone can buy a fractional share in their own name. The company and concept was named Jar, as in a savings jar that families collect their small change in. While a 10,000 rupee purchase of gold might be a big stretch for many people, incremental purchases of as little as 10 rupees can have a cumulative effect and over time can become substantial. “Jar is the custodian of the gold, but you are the owner. You can buy as much as you want or as little as you want, and you can also sell it whenever you want and Jar makes a profit by taking a fraction of the spread. So that was the basic idea but it took us a couple of iterations to get it fully formed.”

One of the rather charming and surprising aspects of this particular Fintech is that it actually does sell gold jewellery – it makes sense seeing as one of the drivers of families saving for weddings is to have enough money to buy beautiful earrings!

Content and yet more content

Prashant points broadly to two types of people in India – those who are well-educated and computer literate, and the vast majority who have had more basic schooling and whose access to the digital world is almost entirely through mobile applications. For them the internet is completely mediated by the smartphone. There is also less access to financial and banking applications, in that English is frequently the main language used for these. Furthermore that financial language is often specialized and rarefied, using terms that are not in general use. Such factors mean that there is widespread reticence about getting involved in financial issues and even in understanding the interfaces enabling this. “People may not have a proper house but they do have smartphones and they consume content like anything,” Prashant explains. “The appetite to consume content is insatiable! And so there are people who are essentially semi-literate, who have access to the cheapest devices and the cheapest bandwidth, and who consume a lot of content, especially video. And the interesting thing – which we often miss about video – is that it doesn’t require literacy. So on the consumption side there is no literacy barrier. And on the creation side, input is by voice. The whole interaction and experience of the internet requires very little literacy, and you don’t even have to type much.”

Getting sticky with it

The characteristics of many Jar customers seem somewhat similar to BharatPe’s, where Saurav Ghosh’s Data Science team plays a central role in shaping intelligence across every stage of the merchant journey. At the BharatPe onboarding stage, advanced models are deployed to conduct fraud checks, ensuring that only quality customers enter the system. As merchants continue to engage with the platform, data-driven models assess product propensity, affinity and engagement, enabling the company to better align services with merchant needs.

In parallel, behavioral models track repayment patterns, evaluate eligibility for working capital and identify merchants who may require top-up facilities. Another critical focus area is merchant ‘stickiness’ in a Fintech landscape where players such as PhonePe and Paytm are competing for the same base. By predicting which merchants are likely to remain loyal and which may switch, the team empowers proactive retention strategies. As Saurav notes, retaining customers is far more cost-effective than acquiring new ones – making predictive retention a cornerstone of long-term growth.

The holistic view

BharatPe approaches the merchant lifecycle holistically, from onboarding and cross-selling to upselling and retention. The objective is to estimate each merchant’s lifetime value and determine the right timing and approach for monetization. Underwriting models are regularly back-tested using AI and Machine Learning, ensuring that risk assessments remain dynamic. Given that many merchants have limited credit histories, robust portfolio management becomes essential, with close monitoring of delinquencies, losses and repayment behavior. “In lending, disbursing the loan is the easier part – it’s recovering it that presents the challenge,” Saurav emphasizes. Technology is used to flag early warning signs of repayment stress, complemented by a measured collections approach that begins with gentle reminders before resorting to field visits if necessary. However a growing concern which needs to be monitored is over-leverage – when merchants simultaneously seek loans from multiple lenders. 

Model monitoring and governance

These models are not static. As customer behaviors shift and external factors come into play – ranging from macroeconomic changes to individual financial circumstances – model accuracy can decline. This challenge, known as model drift, demands continuous vigilance. Models must be monitored, back-tested and recalibrated to ensure they remain reliable in dynamic, real-world conditions.

To counteract drift, BharatPe employs early warning indicators and regularly adjusts key parameters, enabling models to adapt seamlessly while preserving accuracy.

“Once the model is in place, you let it run,” says Saurav, “but with constant monitoring and timely interventions.”

Demand and supply engines

Let’s return to Abhinav Gupta at a different end of the customer profile and scale. IndMoney was off to a rocket-fuelled start, but in September ’23 Abhinav moved over to another radical new Fintech. This is IncredWealth, with the stated aim: ‘To be a premium Wealth Management company for Indians globally. InCred Group has seen immense success in a short timespan owing to its custom offerings in the lending and capital market space. With InCred Wealth, our focus is to bring international standards of wealth and investment solutions while being firmly rooted in our local Indian ethos.’ 

So Abhinav was in effect returning to his banking roots as a private banker and wealth manager. “IncredWealth is fully focused on the HNI segment – the high net worth individual category. IndMoney is a demand engine, but IncredWealth is a supply engine with a lot of in-house products. We have our own private equity, private credit in-house investment banking team and various other alternate investment products. This brings an exclusivity to serve HNI and Ultra-HNI clients through family offices in Singapore, Dubai and London. We have a full team at these places, as well as Mumbai of course.”

The ability to scale

The Incred Group itself was founded in 2016 by veteran banker Bhupinder Singh, as an NBFC – a non-banking financial company – based on lending. 2019 saw the formation of IncredCapital with four verticals, one of which is IncredWealth where Abhinav currently resides. “We have our investment banking team, we have the institutional broking team, and we have the offshore wealth management team. So put together we have been able to grow really fast in the last five years. And broadly the model is applicable to anywhere. Micro markets are different, so you have to adapt to those situations. But in terms of technology, in terms of product offerings, it can be scaled, and similar businesses can be built depending on local regulatory guidelines and ecosystems.” 

Regulation and freedom

Ah yes, the Regulator! Wherever there is banking or Fintech, there’s a Regulator. Abhinav says that the regulatory authorities in India have been quite friendly over the past three or four years. “Broadly, the regulator is trying to promote the Fintech industry in India and in some sections such as unsecured lending or futures and options, where there was some froth the regulator has had to play a role in stopping activities which are not in the interests of Indian consumers. But generally the regulator is facilitating Fintech and also stopping activities which are not good for stakeholders.” 

He goes on to describe activity in Gift City in Gujarat, north-west India, billed as, ‘The global financial and technology hub that offers single-window clearances and approvals under one umbrella. It provides a plug-and-play infrastructure for businesses seeking unparalleled accelerated growth.’ 

“Gift City is being developed on the lines of Singapore and Hong Kong,” Abhinav explains. “Basically, it’s an offshore jurisdiction, so Indian domestic regulations are not applied there. In Gift City, there are now about 15–20 global banks, so you can do all transactions in dollars and don’t have to comply with the local capital account convertibility rules. It’s a completely tax free zone and you don’t have to pay any capital gains here on investments. Gift City is designed to attract foreign capital because we want to develop it as an alternative to Mauritius, Singapore or Hong Kong, where the complete ecosystem allows you to function very smoothly, as per global benchmarks. So a lot of innovation is happening and funds are coming in. I’m sure in the next four to five years it will be a very bright spot.” 

Making it intuitive

Back with Prashant Singh and the advent of Jar: With 350 to 400 million people already using electronic payment, the thinking was: Make savings in the form of gold, designed from the start as an intuitive experience that was approachable. So that’s a Fintech which is a digital piggy bank for a very wide section of the population to use, by saving what are sometimes tiny amounts of the secure and attractive option of gold. It’s easy to understand the concept, it’s popular, and doesn’t require any special financial knowledge or terms. But to return to the gap noticed by Jar’s founders at the outset of the business and that disinterest of banks to direct products towards a huge section of the population: Looking at their own data, Jar noticed the tendency of owners to sell their share of gold towards the end of each month, because that’s when they were running low on funds to pay for groceries, emergencies or other everyday expenses. After all, those are the times when you break into your piggy bank. 

An uncatered-for market

It would be nice if people could go to a bank for a loan to cover such eventualities, but because of many customers’ limited financial literacy, coupled with relatively low income levels, most banks are unwilling to go there. Jar saw a different way, as Prashant explains: “If someone has been saving 50 rupees-worth of gold every day for the past year, we can see how disciplined they are. So if today, they need a 2,000 rupee loan, we’re fairly comfortable with that and can underwrite the loan: ‘Here’s the thing – take 2,000 rupees from us. Don’t sell your gold because we will collateralize it. You spend your 2,000 rupees on whatever you need to cover. Then you get your salary on the first of the month, return the 2,000 and your gold is back again.’ So we have an asset for the person.”

The lending model emerged organically, founded on the ability to talk to users in a language they could understand. With this base model well-established, Prashant says that Jar now wants to be everywhere that money is changing hands. It’s a giant market that the Indian government is anticipating will cover 500 million people using online digital payments by 2027. 

Signs and symbols

It’s also interesting to note how a common iconography is needed as the non-literate but increasingly savvy app-using population grows. Across many applications such as Google and Microsoft there is a commonly understood use of symbols and shortcuts, but Prashant points out that this is not the case with Indian financial apps. What does ‘withdrawal’ look like, or ‘deposit’? Even the western idea that ‘bank’ is represented by an icon of a Grecian-style portico with pillars doesn’t necessarily convey a lot in India. With hundreds of millions of users, there can be many interpretations, and as Prashant observes, “Sometimes you realize that you have no idea how your services are being consumed. The user always surprises you!”

There are some 400 dialects in India alongside the main business languages of Hindi and English, and in some cases even common concepts such as ‘USB’ or ‘homepage’ have no direct translation. The drive is therefore for the localization of language – specifically in terms of documentation – rather than the massive complexity of making everything available to everyone in a multiplicity of languages. That’s another reason for ever-greater clarity around the use of icons. Each new design and feature that Jar develops is prototyped and tested by a dedicated in-house research team to assess how real consumers use the products in the real world. 

Reject Inference: Learning from the ‘invisible’ applicants

Real products in the real world bring us to the concept of Reject Inference, as explained by Data Scientist Saurav Ghosh of BharatPe. Scorecards are usually built on approved and disbursed customers, since their repayment behavior can be observed. The drawback is that models trained only on this visible, lower-risk group may develop a bias, becoming more cautious than necessary and overlooking applicants who could in fact be creditworthy.

Reject Inference is a way to address this limitation. Instead of leaving rejected applicants out of the picture, their data is also analyzed. This is especially the case with the policy rejects: those declined because of internal thresholds rather than absolute disqualifiers such as age limits or severe bureau issues. With the help of bureau data, it becomes possible to see how these rejected applicants performed if they later borrowed elsewhere, and to use that information to estimate likely outcomes for the rejected pool.

This process is not just a statistical fix. It helps models remain relevant when conditions change. For example, during the pandemic many lenders applied stricter criteria, leading to a larger-than-usual share of rejections. If models were trained only on this narrow period, they would continue to reflect that conservatism even after conditions improved. By including insights from rejected applications, the models can reduce this bias and better reflect the broader credit environment. The end result is more balanced underwriting, avoiding the extremes of being too lenient or too restrictive. This reality directly shapes the company’s approach to UX and UI, and for Saurav simplicity is the ‘North Star.’ From onboarding onwards, every interaction is designed to require as few clicks as possible.

On the ground: connecting with India’s merchants

Saurav’s field visits to merchants offered him a unique window into the heartbeat of India’s economy. Observing merchants in person has helped him understand both pre-sales and post-sales operations, and informed practical improvements that could be implemented back at the office. More importantly, these visits helped infuse empathy into the data. Millions of hardworking hawkers, shopkeepers, and small merchants operate with minimal financial support, and seeing their day-to-day realities made the impact of BharatPe’s work tangible.

“I visited multiple merchants – local shops selling clothes, food, people earning their livelihood,” Saurav recalls. “Seeing their joy and knowing that our work helps them grow is incredibly rewarding. It’s not just Merchant ID 1234; it’s a person running a business, living our values. Experiencing that connection firsthand is truly refreshing.”

The Fintech scene

So, viewed from the mega business of international wealth and the booming new Gift City, what is Abhinav Gupta’s take on Indian Fintech in general? He says that while currently India is a very big market, equity penetration is less than 5%. Per capita income averages $2,500 US a year, but is anticipated to grow to $10,000 US in the next twenty years. When that happens he predicts a lot of growth in the financial space – especially the financial services sector – including wealth management, insurance, stocks, broking, and lending. “Overall, the Fintech ecosystem is full of opportunity … but I think India needs a lot of capital.”

One of the big changes in India for the population at large has been the rapid uptake of mobile, especially since the pandemic. Phones and data costs are very low in India and large proportions of the population can now access online-broking Fintech apps and investment platforms. “People can now access and do transactions, and because of educational content on YouTube, they can understand many things. That’s helped dramatically in terms of bringing people into the capital market.”

Fair enough, but we’re still talking about higher-earning, relatively affluent members of Indian society. What’s the word on the street about Fintech? Abhinav says that in most cities, people are generally familiar with the concepts, whereas in rural areas there is less understanding. However, he smiles as he recounts that even beggars are now using India’s cashless Unified Payments Interface to accept donations. That’s a measure of how far the whole digital payment framework has come, in a very short time.

Growth and incentives

Income is heading upwards, the adoption of technology is enthusiastic and widespread – so what does this suggest for the future of IndMoney and IncredWealth? Abhinav says that with the government doubling tax-free limits, people will have increased impetus to engage in a more formalized manner with banks, and then capital markets and other revenues. Now is an inflection point, growing out of the pandemic, and the next ten years will see growth … and then more growth. In the equity segment alone the current 5% penetration indicates massive scope. He views the future positively: “India has a great environment for business growth. There is political stability, and people are optimistic about the future. Our Prime Minister says that we have to become a developed country within twenty years, and we have the aspiration to be world class. People are very ambitious and hardworking and want to really explore the best of things.”

When not navigating international high finance, Abhinav is an avid reader, writer and poet, and makes the connection between art and finance very clear: “If people don’t have the basics, they can’t think about poetry or all the good things of life. Once you get increased prosperity, then those things come. By creating prosperity we encourage culture.”

The young country

“There are many people who really, really believe in the Indian story,” Prashant Singh shares his own sense of optimism. “Whenever the market presents you with a massive opportunity – be it Fintech, video, education, anything – it attracts two kinds of people: the missionaries and mercenaries.” Of course the Jar team are the former, and Prashant shares his obvious pride in their achievements and that of Indians at large: “We are one of the biggest countries in terms of population, and most of us are experiencing the formal economy for the first time. India has no history of access to formal lending, but once everyone has that access, our power will increase. By bringing everyone into the financial fold I think we are at the onset of something huge. There will be false starts, there will be difficulties, there will be failures, but this is a once-in-a-lifetime opportunity.” Prashant tells his personal story, “I grew up in a household without a TV, and my family’s evening dinner used to happen around the transistor radio. Today – and I’m not that old a man! – there are kids in my team who don’t even know what a radio is. One employee didn’t know how to use a rotary phone and he’s a pretty sharp guy! But that’s the pace of change we are dealing with in India. We are lucky to be in the country at this time.”

A fire-level opportunity

The concept of dropping loose change into a savings jar is simple to grasp, but how will AI play a part in India’s Fintech future? We’re back to the view of a missionary because Prashant’s own background is as a founder of an AI-based company which he sold to India’s largest payment processor, Paytm, back in 2015. “How big is the opportunity for AI? Is it internet-scale? Is it a smartphone-level opportunity? Is it an electricity-level opportunity?” He pauses dramatically, “I believe it’s a fire-level opportunity. It’s as fundamental as humanity discovering fire, and I don’t say that lightly. Understanding and building intelligence will unlock so much of value, and we are finally moving from an age of scarcity towards an age of abundance. It’s huge. We’re lucky to have a ringside view and to be participating in it. Fintech is all about finance, and money is all about the exchange of values. For the first time in history we have all the data, and the insights you can get out of it mostly depends on how you slice and dice that data. At a very fundamental level, these algorithms give the power to understand users in a very unique way.”
And from that heady call to action, a final reminder from Prashant that apps should never be boring: Jar has proved that there is no separation between Fintech and ‘normal life’ where cultural and financial trends meet and mix.