I think where AI excels is in trying to emulate what could be a human interaction.

And, by definition, what we do, and specifically do for mortgages, is dominated by human interaction. Across the world - so we're talking not only about emerging markets, but developed markets as well.

You have mortgage brokers that talk to customers. They do simulations. They explain what the bank offers that they can get approved for, and so on and so forth.

So I think the real need here was how do you bring that human experience into a place where you have thousands of islands and very remote locations and where the people buying properties are not necessarily concentrated around the large cities.

We're talking in India of about over 50 million customers. So numbers are huge.

And this is only in India we're talking about. But even here, in Canada, also, you have some microfinance, a lot of people are involved in this. And there's some very large institutions in India, which does this.

85% of folks have bank accounts, but only about 20% of them actually use the bank (for borrowing). They rely on other forums to meet their financing needs. These guys don't have access to credit.

As a startup, when we launched back in 2018, the lending market in Sri Lanka was $5 billion - and what we understood this from this 5 billion requirement, almost 40% relied on loan sharks and individuals who have very dubious practices of charging excessive interest rates, as well as very unpleasant collection methods.

And coming from a tech and a finance background, we realised there might be a potential for us to provide credit underwriting through a blockchain solution.

And then the credit scoring engine started taking shape. And over, you know, Covid, after Covid, we started bringing defaults down from 50 to 40, 40 to 30, 30 to 20, 20 to 15. And then, you know, the tougher bit came because it was not only about the credit scoring, it had to be a lot of engineering, then it's about the experience of the consumer, how is the product structured, you know, the first interaction of the consumer with the company till after he or she has repaid, everything matters. How the lead generation happened, how is the customer support interacting with the customer, or what has been experienced in app, what is experienced at the point of repayment, you know.

We've seen many times if the customer faces challenges in repaying whether or not it was our issue or an issue at the wallet side, the customers could turn rogue.

There was so much to it that we learned over time. And you know, when we actually closed our lending book pre-acquisition, the latest cohort actually close it under three per cent default