L&T Finance Ltd

Good evening, everybody. I would like to thank all of you for taking time off from your busy schedule and coming and attending this Digital Day of L&T Finance. As you are aware, during Karthik's introductory message that we are celebrating our 30 years, our pearl anniversary of L&T Finance, we are founded on the 22nd of November, 1994. So it's been an eventful 30 years and to commemorate this, we are having our Digital Investor Day today and tomorrow we are having our main marquee AI conference, which is called RAISE and which is about applicability of AI into BFSI and all of you must have got invites to it, as well as I would urge you to take the entry passes for tomorrow's event because it's going to be a very exciting day tomorrow.

I joined L&T Finance about 18 months back and for those some of you might know me, but for those of you who do not know me, probably I'll just give a brief background about myself. I spent the last 14 years of my career in ICICI Bank, where I used to manage unsecured assets, cards, payments, parts of digital banking, specifically the connected banking and API banking, the student ecosystem, the merchant ecosystem, as well as the digital acquisition piece for liabilities. Prior to that, I spent about 13 years equally spent between Citibank and Deutsche Bank. I joined L&T Finance in July last year and it's been my 18 months running now in L&T Finance. You know, it's been a very, very eventful 18 months and what we have tried to do in 18 months was to sort of turbocharge the Lakshya 2026 goals that L&T Finance had already embarked upon and sort of move the needle forward in the digital transformation of L&T Finance.

Now obviously, many of you are in the room today because, you know, many of you are probably wanting to seek answers to some of the sort of top lines of our MFI business, but what I would like to tell you is that, we will give you top lines and data about our MFI business as well, but L&T Finance is not only about only MFI business.

In fact MFI is only about, you know, one-fourth of our total AUMs. A lot of things are not known about L&T Finance and I thought that, I will sort of lay out some of them before you. Obviously, we complete 30 years. We have a very large customer franchise. We have a 2.5 crore customer franchise. Out of it, 1.1 crore customer franchise is in rural. We are one of the very few NBFCs who have a balanced customer profile between rural and urban. Our customers are divided between 50:50 rural and urban.

If you look at our employee strength, we are about 34,000 employees and in about 13,200 distribution points and a Pan-India presence of over more than 2 lakh villages and about 100 cities. In terms of our urban business, our urban business is almost Rs. 42,000 crores with about 100% digital onboarding.

We have best-in-class ESG rating, CDP A-, and we are AAA rated by all the four rating agencies. If you look at our other sort of parameters, we are one of the largest JLG financiers in the rural space. We are the third largest in the country. We are the leading tractor financier in the country and our onboarding systems, 100% of our onboarding systems is paperless, completely paperless.

Now, one of the things, you know, as we stand in 2024, I am reminded of a quote by Charles Dickens, it was the best of times, it was the worst of times. It is in the tale of two cities and it is the best of times, primarily because the world is at the cusp of the fourth industrial revolution. The world is at the cusp of massive changes in technology driven by artificial intelligence and machine learning. The space race is again hotting up. People are talking about putting man on the moon and then man on Mars in the next 10 years. And the fact is that, we are seeing that the Global South or the developing countries are picking up growth. However, it is the worst of times primarily because in this environment, we have two wars raging, one in Europe and one in Middle East. And overall closer to the country there are certain sort of pockets of doom and gloom, which are visible in the asset space. So that is why I term it the best of times and the worst of times.

But obviously, we focus on the best of times and this is the time where we have decided to kick forward our digital transformation and take our transformation agenda much, much faster ahead. If you see, we started our Lakshya journey on the Quarter Four of FY 2022 and now as we stand in Quarter Two of FY25, let's look at what we have achieved. We have transitioned from a wholesale dominant franchise to a retail diversified NBFC. The retailization currently is at 96%. Our retail book grew by about 97% and we realigned our strategy from a product focused company to a customer focused company. We posted the highest ever PAT of Rs. 696 crores last quarter, which is up by 104% from what we were in quarter four of FY22 and if you look at our ROA at about 2.6% and ROE about 11.65%. Our credit cost has trended downwards. Last quarter it was about 2.59% and our asset three -- asset quality in terms of GS3 and NS3 remains at about 3.19% and about 0.96%. In the interim, we have added about 5 million new customers. Obviously, the customer franchise has grown by 26% on the back of a robust distribution network about 28,000 partner touch points. And last, but not the least, this is one of the most important part. Digital led the core of this transition. We launched our PLANET App about three years back and currently our PLANET App, which is in PLANET 2.0, PLANET 3.0 will be launched soon. Currently, we are seeing about 13 million customers regularly using it and we transitioned from a disjointed incremental to an integrated modular product technology approach. Big words, difficult sentence, but I'll explain to you what that means in the next couple of slides.

Over the next 18 months -- over the previous 18 months, we have picked up speed in execution and we have been executing at a reasonably fast clip. So, in the last 18 months, we have sort of grown the retail book by about 38%. And if you look at our disbursements, actually disbursements have grown by about 35% from about 11,000 crores at the end of Quarter One FY24 to about 15,000 crores last quarter. Our rural business finance business has grown by about 34% in terms of book size. In spite of a tough operating environment, last year our tractor business grew by about 10%. Our two-wheeler business grew about 38%, personal loans. We actually slowed

down personal loans last year for about two quarters. We wanted to make sure that the digital journeys are retooled. Our risk metrics are far, far more strengthened and then we have again started growing our personal loans business. Sanjay Garyali, our Chief Executive for Urban Finance will cover it in his presentation. Home loans and LAP has grown in about 54% and SME has also grown at a very high rate albeit on a slow base.

If I look at the other work that has got done, we have transitioned from a Silo structure to a Matrix structure. We have divided the organization now into four regions, distribution regions. We got four very seasoned distribution professionals as Regional Business Heads and we have moved the organization to a matrix structure for a far more granular distribution focus. We have been recognized as a great place to work. We have built a 3-Tier compliance structure. Our number of people in compliance last year at the same time, probably were eight people.

Right now we are about 3x that number. We have put together a regional compliance structure. We put together a very strong compliance testing framework to make sure that the business that we do is done within all ambit of regulation. And last but not the least, I will spend some time on the technology part and though it's mentioned here that you know we launched Project Cyclops and AI-driven underwritten engine, in fact this will be explained in detail during the presentation by our Chief Digital Officer and our Chief AI and Data Officer and you will have a fair understanding of this at the end of today's presentation. We institutionalized NPS tracking for all lines of business starting of November 23 and our Chief Operating Officer, Raju Dodti will cover the details of that in his presentation.

In terms of technology, I will spend a lot of time on technology in the next slide, but the fact is that the focus has been on building omni-channel digital architecture which means you know you build solutions or you build productized solutions which can operate on any channels or operate on any products. Use AI and ML technologies at the core and you will see how we are using at the core in the next slide. And last but not the least, build in- house application engineering capabilities. We have more than 300 application engineers right now divided into risk engineering, application engineering and product engineering. We have put in a state-of-the-art technology center in Bangalore. Our second technology center is coming in Mahape in Mumbai and we already have 300 of those engineers on board. In fact, Cyclops which is our first three-dimensional AI engine probably for the industry, create engine for the industry was built by our application engineers flat within a period of four months. And last but not the least, we have built a micro service based API stack which obviously is currently live on these two particular partners. We are going to announce another large partner tomorrow. So, if you are in the session tomorrow, you will see which large partner we are announcing tomorrow. And last but not the least, we have ramped up our brand visibility. We have launched a multi-channel brand marketing effort. You saw the ad, so I'll not spend too much time on it. And obviously Captain Bumrah has done boom to Australia today.

So this is, I will actually quickly click through it so that you get the full picture. This is our actually unified technology stack which we are imagining for tomorrow. You have to understand one thing that technology right now is moving at a very, very fast pace. Traditional methods of delivering technology or traditional methods of delivering financial services to technology is probably dead. Especially nothing more is more dead than your credit underwriting. And you will see we are using one word very effectively in this; that word is intelligence. Why do we use the word intelligence? Because the current tools which are available in terms of compute power, in terms of data availability and in terms of tools, whether be it in machine learning or through the AI tools which are currently available to make much more meaningful inferences out of the data and the structures about every customer which is available, and give inferences either in terms of credit decisioning, in terms of portfolio management or in terms of service. So that is why we are transitioning to an architecture which we call an intelligence-based architecture.

And you will see in the presentations of our Chief Digital Officer, as well as Chief AI and Data Officer, how we are unpeeling these intelligence layers and building products around these intelligence layers.

I will take a little more time on credit intelligence. Traditional methods of underwriting I said is dead and I will repeat it again. Traditional methods of underwriting is dead because traditionally we used to take the customer's detail, we used to take the customer's income, we used to go to the Bureau and then used to underwrite. But the fact is that the data which is there in the Bureau, though it provides a rich history of the behavior of the customer,

it takes a little bit of lag data. Customer data is available on tap, on demand and in huge volumes, especially in India where during Diwali we almost had 800 million UPI transactions. A huge amounts of data are flowing through, both in terms of payments, in terms of micro-geography, or in terms of social media, or external marketplaces, especially e-commerce data. Now if you are able to triangulate all this data, which I call trust signals or micro- geography data you tend to get up a far, far more richer understanding of the customer, which is what we mentioned as credit intelligence block out here. In fact, one of our major developments was Cyclops, which is the sort of the first product in the credit intelligence block, which we have built and we have operationalized in the month of June this year. And we are steadily scaling it up with very encouraging results and Debarag and Ramesh will sort of share some of the details and the resiliency of Cyclops in their presentation. Again, I'll click through all this. This is sort of the sort of the unpeeling of the various intelligence layer and what are the product boxes that lie in it. One of the things which I said initially was to move from an incremental disjointed approach to a product approach. That means when we take a solution, we build a solution like a product, that solution becomes a product box which is omni-channel, omni-product. That means, if I build something from two-wheeler, that particular skill of the products box that I have built for two-wheeler, I can actually pick it up and put it in for SME. I can actually pick it up and put it in for mortgage. What it does is that, it gives us this modular approach to product building and makes product rollouts much, much more faster. Like, for example, in portfolio intelligence, when you do portfolio management, most portfolio management in the country is reactive. That means, you really react after the horse is bolted. You do not have proactive portfolio management because on large portfolios, when you have 25 million customers, when you have 30 million customers, when you have 50 million customers, you cannot monitor customers individually. Current processes or technology of monitoring is basically you take cuts and then you try to make inferences based on cuts. But what this engine, for example, this is our next build which has started, which Debarag will spend time on, who is our Chief AI and Data Officer, is Project Nostradamus. It is basically a very high-end automated portfolio management engine that the engineering team is trying to build, which moves L&T Finance from a reactive portfolio management paradigm to a proactive portfolio management paradigm. If you see, in the consent layer, we are already live on all these tools. These presentations are available and uploaded. You know, account aggregator is something that we have been focusing big time on. Right now, our account aggregator success rates are 91%. That means for every 100 pulls that we do on account aggregator, we get 91% account aggregator success, and Ramesh will cover some of those metrics in his presentation. If you look at ability to understand the customer, customer lifestyle index is an AI-based tool that we have built. This Debarag will cover in his presentation and we filed our first patent on customer lifestyle index. In fact, what my guidance to the team and what our guidance to the team is that, make 100 moonshots. We are not afraid of failure.

We will pick up 100 projects. We will make moonshots against these 100 projects. We will try. If we fail, we will fail fast. But the fact is that out of these 100 moonshots that we take, even if 15 or 20 succeed, they will lead to a transformational outcome of the organization. That is why what we want to go forward with this particular organization. As I said, you know, in the ability to underwrite and the credit indigenous layer, the greens are the ones that have been sort of done yet. For example, decision through additional trust signals such as satellite data and digital footprint. We are using it in the beta that we are going to roll out in the month of December in underwriting our tractor business and Ramesh and Debarag and as well as Asheesh will also speak about it during their presentations.

Now, what will all this lead to? Obviously, this will lead to scale up of business at a low cost of acquisition and obviously lead to higher productivities. It will lead to stable and predictable credit costs. This is something that we do not want to compromise on. In fact, there is a joke internally where I say that all our middle name especially is credit. That means finally credit, you have to deliver credit with certainty. You have to use all tools to deliver credit with certainty. And if you are actually stable in this, you run a stable organization. Otherwise, you run a sinusoidal business approach which does not serve anybody right. Obviously, our focus is best in class customer satisfaction.

We talked about service intelligence. We will give you some examples of service intelligence in the next presentations. As I said, the modular approach allows us faster rollout of new product and stabilization. So the modular approach is allowing us to do products much, much faster. I talked about one particular product which has rolled out; especially Cyclops was rolled out in four months. Most of our products are seeing much faster turnarounds. And last but not the least, obviously, our expansion will be geo expansion, you know there are certain

Overall, we are an organization in the middle of enormous transformation. We are an organization in which we have actually sort of trying to push the needle on transformation on all lines. You know, on people transformation, we have hired some of the best people across the world. You know, our Chief AI and Data Officer, I searched for them from all over the globe and got them from Silicon Valley. Our Marketing Head, she came from Pidilite and managing Fevicol, a marquee brand in the country. Our Business Heads, whomever we have hired new, they have had significant sort of track record. So in a way, we have sort of tried to position the organization for the leap forward. And in today's presentations, which my colleagues will take you through, I hope we will give you a sense of how we are taking the leap forward, apart from giving you some business data for which all of you are interested.

Thank you so much. We have demo stalls of some of this which have been sort of made and kept in the hall next to, this particular hall. And once I am through, we are through with the presentation, we would like you to join us in the demo stalls and see for yourselves some of the solutions, how they are operating. Thank you so much. And I will see you around after the presentation is over. Now I will hand over to Ramesh. Ramesh Aithal, our Chief Digital Officer, to take you through the digital transformation part of our journey.

Thank you. Thank you, Sudipta. Welcome, everyone. I'm here, obviously to take you through how technology is transforming the BFSI space and how L&T Finance is also kind of fronting it. All right, before I start off, perhaps a brief about myself. I've been in the industry for over two decades, primarily worked early on in my career with Fintechs like Citibank and Goldman Sachs, where I stayed for a very, very long time, more than a decade, primarily working on products, building trading platforms for trade derivatives, mortgage derivatives, futures, options, interest rate products and structured products. Built that for a long time and then moved on to manage the risk platforms for Goldman Sachs; primarily for Asia, market risk products. Again, you know, shocking stress, the portfolios and figuring out what implications are there for the portfolio at Goldman Sachs. Moving on from there, after like I said, over a decade, started working at a mid-level Silicon startup, Silicon Valley startup that was primarily working on Payroll and HR software and led the teams for India, and then moved to a more Deep Tech Valley company named Elastic, where we built search engines for Fortune 500 companies. Around 10 months ago, I ended up taking this role at L&T Finance and here I am.

All right. Before we move on, I want to kind of perhaps take you through some trends across in the BFSI industry.

No technology presentation would perhaps be appropriate without saying our pronouns to Generative AI. In the last couple of years, since it's come onto the mainstream, it's moved from primarily an experimental project for the most part, and in certain pockets of technology domain to the mainstream. And while it has moved to the mainstream, now we are beginning to see applications in the real world employing this technology.

A few key features of that, one of the things that came out in a recent research report by a VC firm in the Valley is that, in Generative AI, just over the last year alone, the spend has gone up six times, from $2.5 billion to around $13 billion. And this is, again, keeping aside all the huge spend that the big tech is doing anyway, primarily on those foundational platforms. Aside from that, what is noticeable is a shift from broad-based models that we are all used to, like the ChatGPTs of the world to more vertically integrated models that are being used in specific areas in finance, sales, marketing and so on. I think the future is here, especially in this field and we are witnessing it first-hand. Moving on from that, another hot topic at the moment is about hyper-personalization. I mean, we are in the age of instant gratification. Whether it be your hyper-local groceries or how you consume your content, everything is getting personalized. Very soon, we'll be able to kind of offer you a lower, interest rate mortgages in the development next to your office, based on the commute times that you are spending in Mumbai. We're not there yet, but we'll get there. The next development is about the open lending ecosystem, the India Stack that Sudipta also referred to earlier. This is an area where there is a complex interplay of market participants, the government, NBFCs, banks, all playing their own part and building a very competitive flat ecosystem. While this is great and there are business implications, it's very, very interesting because from a technology standpoint, it dramatically increases the complexity of the systems that we build. Just as an example, in our supply chain finance business, recently, we built an integration and we onboarded a new system and there were more than 30 plus external integrations that we had to plug in. That tells you the extent of the complexity, and I'll come more on that in the subsequent slides.

The next part is around Cloud. Cloud is used in 80 % plus of the industry already. In most banks, you have heard of Cloud, but even Cloud architectures are evolving into more managed services, primarily because of the trends around how managing it and trying to bring in more robust data resiliency in the industry.

The last part, of course, is data protection. On the one hand, you have the government with its data privacy laws, the upcoming DPDP Act and on the other hand, you have threat actors that are constantly evolving strategies to disrupt the business. In the middle, we need to maintain our sanity in terms of defending ourselves from these actors. Moving on, this is a slide that you've just seen from Sudipta and I'll do a deep dive into some of these layers where we've been able to impact or in the process of impacting, from a technology standpoint. The first one, of course is the customer consent and verification layer.

Customer consent, obviously, this is critical. Customer consent and verification is integral to any business that we do at our company. It is a cornerstone of trust. A few years ago, a lot of customer information and data was collected on pen and paper, just a couple of years ago as well. With the India Stack, now you have a consent architecture in place. You have a complex interplay of companies, like I mentioned, aggregators, data sources from the government primarily, as well as industry participants, and we have our NBFC kind of consumers of the data. This is leading to a lot more frictionless customer experience in the market. A couple of examples I want to take you through here is primarily, the first one is around what we call as our account aggregator implementation.

Here, we've taken the process of collecting a customer's bank statement and verifying it and using that in order to process a loan to a much, much more early TAT or Turn Around Time from around 15 minutes or days in some cases to seconds, 15 seconds. This is something that's been possible primarily because of the ecosystem and we are a big participant in it. The next example is around how we are able to use, again, customer information to provide them a better way of managing their finances. The PFM tool app is actually going to be launched within our PLANET app in the coming weeks. Going back to take you through some data around how many integrations we've built, there's a sizable number. I'm not going to cite through all of them. The only thing I would want to call out is the metrics on the far right side. We have 100% of onboarding of customers happening on PL digitally via V-KYC. E-sign is 100%. Again, cashless collection in our businesses, 60% excluding some of the rural businesses. Customer onboarding via DigiLocker is 80% and I'm happy to report that this number is likely to go up again further because just last week we received our biometric KYC license as well. All in all, this is something that we are definitely investing heavily into to provide a compliant, convenient, and frictionless customer experience.

The next layer I'll talk about is customer intelligence. What we're trying to do is get deeper into the minds of the customer. How are we doing that? Again, harnessing the ecosystem. 50 plus fintech partnerships already built.

And more on the way. What that enables us to do is collect up to 10,000 data points. 10,000 data points. Humanly, this is not possible happening through the technology that we have. Customized offerings enables us to get early warning signals, propensity to default, and much more. Again, I do have a couple of examples to show you how deep we are getting into how our information gathering as well as our customer onboarding is. Two examples here. The first one is the video personal discussion that I'll share a quick example of right now. This is our friend, our credit manager, Dhananjay Singh, in Bihar talking to a customer about their tractor business and understanding this information in their credit appraisal process. The next one is in the microloan space. Currently, our microloan team wakes up early in the morning, plans their routine to collect money from the customers and fairly long route in remote areas and remote corners of their vicinity, of their jurisdiction, as well as, you know, finding out where the banks are. All this information fed into the system, we're able to provide them the most optimal route to collect the money as well as deposit the money. The next part is about credit intelligence. As you saw in the earlier slides, we had data, we collected data, we have developed insights into the data. How do we use it. That leads us to credit intelligence. Sudipta referred to the product project Cyclops, rather. The story goes that sometime early this year, Sudipta set us a target about building the next state-of-the-art credit underwriting system with higher approval rates, the ability to use alternate data sources, the ability to use AI models, runtime customer evaluation, and also do it in a very, very rapid way with high enough volume as well. And over the course of the next four months from early Jan, we put a multidisciplinary team together across different businesses, as well as technology, analytics, AI, ML, all those teams came together to deliver this completely in-house. I want to take you through the high-level system design of this product. As I referred to earlier, we have data that's coming in from the outside, alternate trust signals, geo-signals, AI, and the bureau. We have internal data coming in from the data lake and data warehouse that we have from the customers. We have our AI models also deployed, and from the front end, we get our applications to give us customer information.

And we have the system that's kind of going to process this data at lightning speed. From an engineering standpoint, this challenge we've undertook kind of perhaps falls into a certain framework, what we would call it as the 4V framework, and that's how it's been designed. It's been designed for Velocity, Veracity, Variety, and Volume. I'll take you through that, starting with Variety.

Today, we know that data is not consistent. It's coming in from a diverse set of sources. We have 16 plus alternate data sources, and we have the ability to handle disparate data sources seamlessly. The next one is Volume. We have massive volume handling capacity. This is built on a microservices-based architecture with the ability to horizontally scale. And today, we have currently benchmarked it for about 200,000 transactions per day. The next one is on Velocity. Sorry. This is about the ability to make in rapid decisions quickly. And to be able to do that, we have employed fast API, a very low latency design, as well as what I call parallel processing. We're able to take 100 hits per second, primarily on most of our APIs. Finally, couple that with the Veracity part of it, which is the ability to deal with incomplete datasets. Like I said, you have 30 integrations in some cases for a simple business, and what that does is you have to rely on data being incomplete, and to be able to deliver information and process even with some of those data sources not being prevalent or being faulty. And that is the ability that we have built with this system. 100% data reliable with no data loss. All-in-all, we've delivered this for our two-wheeler business a few months ago, and farm is coming up very soon, and subsequent businesses will follow after that. Just a snapshot of what we have delivered. We have the ability to process 2,500 data variables and crunch it down to 733 data variables on which we make our decisions. As I said, API, super fast APIs. 4x improved throughput, and finally, this was tested during our recent Dhanteras where we had up to 56,000 hits. The capacity, as I told you, is around 200,000, so we've still not reached there. With other businesses, our hope is that we'll get there very soon, and then we'll scale up again.

The next part is around portfolio intelligence. This is somewhere, again, our tech team is quite excited, as you can see, with delivery of Cyclops, and now we are looking at the next frontier to cross, and that's about portfolio intelligence. This is the project that Sudipta talked about, referred to earlier, which we call as the project Nostradamus. The idea is to build a state-of-the-art automated portfolio management system with predictive risk

management capabilities. We have an initial system blueprint in place. We are happy to share that with you.

Again, the key features will be the ability to push data at scale, to process data, again, at low latency speeds, and to develop the ability to predict and risk model and run your simulations in real time. Finally, obviously, for the users, to provide them with functional views that they can use in day-to-day decision making and decide, you know, how the portfolio is doing. All of this, the team has already been put in place, and it's a completely in-house team, like I mentioned. We have conceptualized the time, the design, we are working on policy and the next steps, and we are planning to roll out in Q2, Q3 of the following year. Pretty excited about this development for us.

The last part, of course, is service intelligence. I mean, most of the earlier parts, layers that we talked about, they're point-in-time things, usually, usually. But service intelligence is something that stays with you for a long time, because customers stay with you throughout the tenure of a loan. And what we've done is, we've, again, massively invested in digitizing the customer experience, so that there is a huge element of personalization that can happen, the huge element of engagement that can happen with the customer. We've increased our digital servicing platforms to seven at the moment, from one. We've gone to 200-plus servicing options completely online, self-service, and that has taken us to, in two years, to 90% digital servicing of our customers. We don't intend to stop there. We continue to invest in automation. Robotic process automation has enabled 80,000 hours of savings in our operations functions, just pure last year alone. We are focusing more on the analytics that we can derive from our speech and call centers, as well as investing further in our flagship D2C offering, which is PLANET.

That takes me to PLANET, which is our offering. Launched a couple of years ago, we had 1.3 crore user downloads right now. This is obviously enabled by around 208 services and 18 business journeys already on the app. Seven engagement journeys, and obviously, in India, we've got to be multilingual. Here are some charts that takes you to the extent of the impact. We have around, the number I want you to look at is the 1.9 million MAUs we have on the app alone. We've done a business of more than Rs. 1,500 crore, around Rs. 500 crores coming from new businesses, Rs. 1,100 crores from cross-sell, and we are already at 4.5 on the app Play Store, and we have the top search results on a lot of the keywords that you would normally employ when you search for a loan.

Evolving on that platform, 1.0 was primarily on correction servicing two years ago. Early this year, we launched 2.0. This is coupled with all the business journeys, and now, within the next two weeks, you'll hear about launch of the 3.0 app that will be a complete revamp on the user experience, a brand new design with a completely revamped backend as well. All of this, again, focused on delivering a consistent user experience across all our platforms. I'll take a short detour from here to tell you how this is possible by taking you through three broad categories of modernization that we're employing. The first one is on productivity. You can't get to where we want to get to without focusing on our productivity of our teams, and three specific call-outs there is on the speed.

We've been able to develop things faster, 20% increase, primarily by using a lot of AI tools. Moving on, agility. By employing agile practices in the industry, we're able to get to 50% more productivity, even with business scenarios changing so rapidly. And finally, a lot of our change management has been streamlined and 40% reduction in incidents, primarily because we are able to develop tools and deploy them for automated testing and deployment Moving on, API stack. I already talked to you. We are focusing on super-fast APIs. Primarily, if you see, we have 500 plus APIs right now, and we are benchmarking to around 30 milliseconds on average response time, and we intend to take that further. What that allows us to do is to reuse code, deploy them quickly in a modular way, and reduce risk of failure because you can plug and play as you need them. Finally, cloud compute. We are going to evolve further from cloud, single cloud dependency that we had earlier to a more hybrid cloud mode. And the reason is, it's primarily because dependency on one cloud or one vendor or one set of rules is always constraining.

As an NBFC, we know we have to evolve and cloud architectures are evolving as well, and so we'll be moving more to a more public private cloud setup that will enable us to deliver on our user experience goals as well as on our scale goals much easier. And on this topic, RBI themselves wanted to get into the act by competing with Google and Microsoft, and I think just last week they announced a new cloud offering that's going to be coming and that will be available to financial industry participants.

Obviously, the next slide is around cybersecurity. There is a whirlpool of insecurities that keep cropping all the time, and if you are not careful enough, you can get caught. And we constantly have to focus on being one step

ahead of our threats. A lot of names here might be familiar to you. What's interesting is something called a supply chain attack, where threat actors identify the weakest link in your supply chain and get a backdoor to enter your systems and take down or cause, let's say, malicious content, software, anything like that, deploy that by coming in through the backdoor. And that's something to watch out for. That's very new in the industry. We've built a Zero Trust Architecture that would enable every device to authenticate itself and register for every transaction that would happen. And we continuously monitor it through human as well as automated AI, advanced AI systems, to enable that we are, like I said, one step ahead of the threats.

Finally, regulatory compliance forces us, because that's the oxygen of the entire industry, to make sure that we are on constant vigil. Finally, like I said, like most folks here are people who love numbers, so I thought I'll leave you with a slide where we talk briefly on the metrics that we've managed to kind of call out, share with you some metrics, not all of them, but a few key ones here, where first one is on the sourcing, where we'll be able to bring down sanction to less than five minutes for our ML, PL, and two-wheeler businesses. Disbursement collections, again, from minutes to seconds. Servicing, 24 lakh rural users on PLANET alone are using our app. Rural users, not urban. Retention, 10 plus offerings, primarily driving engagement. Increasing engagement, more on the app.

These are metrics that I'm really happy, proud of to share with you. Because there have been efforts, sustained efforts made in this direction, and they're trying to bear fruit. 40% increase in cross-sell, 60% increase in D2C disbursements, direct-to-customer disbursements through our app, mostly. 25% increase in offer conversions, and 30% increase in contactibility and search through our advances. To sum it up, we are future-proofing growth with smarter engineering and cutting-edge technology. Thank you. I next call upon Dr. Debarag Banerjee, who's our Chief AI and Data Officer, to take the stage.

Thank you, Ramesh. So we heard of how we are transforming the business of lending, and I'm going to talk about how we are doing it through AI. Briefly about me, I started this journey of AI 30 years ago when I wrote my first paper on Neural Networks published at IEEE back in 1994 when I was doing my undergrad back here in India at IIT Kharagpur. Then I went over to the Valley, did my PhD at Stanford, founded two successful companies, both exited to public companies, one of my few, and served as executive leadership in tech companies that are household names by now. Then my career took a right hand turn where I was part of this journey by Reliance to connect all Indians, 200 million plus, through the launch of what at that time was World’s most advanced 4G network JIO. I stayed over to then bring the first commercial GenAI application, that too in the field of fashion, at Myntra, which is as part of Flipkart now acquired by Walmart. Fast forward to now, I am now heading AI and data right here at L&T Finance, and what excites me here is that I saw India getting connected, highest volume of mobile data with the lowest prices anywhere in the world. On top of it, e-commerce is growing, newer and newer applications every day, and not just e-commerce but e-government. So all of the government paperwork is coming digital, UPI, Aadhaar, and so on and so forth, the India stack.

So what is next? I think this mass of data with the combination of AI as it mature, as I have seen it mature through all these 30 years, is just right to bring credit to the masses and unlock value and make that frictionless, personalized experience with a deep understanding of all Indians. So that's what brings us here at L&T Finance.

Sudipta and Ramesh talked about the pillars, the five pillars of intelligence. I'm going to walk you through them one by one.

Credit intelligence. The first step that we took in that is what we heard about as the project Cyclops. Just like Cyclops the giant, we are looking at data that is not just credit bureau data to understand each and every customers who come and apply for a two-wheeler loan. With their permission, we look at their bank statements, their digital payment signatures, and also geo-intelligence based on their home location addresses as to what kind of affluency is there in the neighborhood. And then with these data correlated with the past performances of our past customers, we can build delinquent models for each of them. For not just credit, but banking information models, payment information models, and also geo-intelligence models. And they produce, as you can see, a

decent level of accuracy. GINI is a commonly used score, as you know, especially in the financial industry, for measuring accuracy of especially classification models. However, what's interesting is when we then combine this information together, we arrive at an accuracy that is far better than even the individual accuracies of the individual models. And that leads us to a much sharper razor. With that razor, we can segment customers from the best to the worst based on what we predict is going to be their delinquency rates. Now, what do we do with the segments? So for the best segments, we give the best offers. Segment one, we give 97.5% LTF on an average so that they only have to pay 2.5% in down payment. Very good interest rates. But then for the less good segments and the worst segments of all, we increase the interest rates and decrease the LTVs. The result is that when we look at the actual uptake of these offers, the ABNDs, which is the Approved But Not Disbursed Rate, we see the best segments having the most uptake because they have the lowest ABND. But it goes up as we go up the stack in terms of the worthiness of the customers. Now, what does that do? It means we are bringing in more of the good customers and less of the worse. So you can actually see that effect in our customer distributions. The blue is how our customer distribution was before we launched Cyclops. Now, because more of the better customers are coming to us and less of the worst customers, we see a definite distortion in this distribution towards the better customers. Now, not surprisingly, that means we are bringing down the delinquency rates because better customers bring better reliability. This is already evidenced by a 64% fall in net non-starters. Now, you may wonder, is this better performance coming at the cost of not making enough loans? Actually, it's the exact opposite.

So, we are right now running essentially an A-B experiment where we have two legs, the non-Cyclops leg in yellow and the Cyclops leg in gray. And over time, we have ramped up the percentage of loans that are going through the Cyclops engine. As you can see, until the launch of Cyclops, they were in lockstep. But after that, a divergence happened. We are actually bringing in far more customers through Cyclops than is happening with the non-Cyclops leg, which is why we are, in a measured pace, increasing the number of people coming in through Cyclops. So this is proof that all of this alternate data to understand customers actually work. But we are not just using Cyclops in two wheelers. We are taking it to rural India as well. The first step is farm loans, tractor loans.

We are one of the bigger tractor lenders in India and part of the world. The difficulty with tractor loans is that it is very hard, apparently, to predict whether a particular farmer using a particular tractor in a particular land is going to have a good outcome, good financial outcome, good income, and so on and so forth or not. And we have actually been able to crack this puzzle, not only by using all of those other information of two wheelers that I described, but in addition, in this case, we are bringing in data from the meteorological offices, the weather data, as well as satellite information. So from that, we can actually enhance that model and bring in greater accuracy in predicting, given a particular type of soil, given a particular depth of soil, given a particular temperature range or rainfall range, whatever may be applicable for that particular soil, whether that tractor loan is going to have a higher or a lower probability of going into default. These sets of models are going into production as actual digital underwriting models by the end of this year.

Coming to customer intelligence, one of the harder problems in credit is understanding whether a particular customer is going to go prime or not before they have gone prime, according to Bureau records. Bureau records, as you all know, are lagging indicators. So with all this other data at our fingertips, is it possible for us to catch that customer who is going to be a good customer early or not? And we believe the answer is yes. To give you some thought about it, about what are the, you know, kind of ways you can think that might be possible, imagine a young fellow coming from a good neighborhood, taking a tractor, a two-wheeler loan. We see that he is using a good phone, he is dining at the best places in town, and he is going to a good college. He is probably set well enough in life that he continues, that although his Bureau record at that point is pretty thin, he will probably go on a journey that will eventually make him one of the prime customers who will eventually end up buying an expensive house, expensive furnitures, good vacations that he will take on credit cards and personal loans that he will not only take but pay off pretty early and so on and so forth. All of those are fingerprints that are available from all of these alternative sources of data. So, we are doubling up on understanding customers above and beyond just at the time of disbursement and following their journey and understanding which ones will become prime. Not only can we do this for urban India, but even in rural India, imagine a family that where the lady of the house starts off with a microfinance loan, but then it turns out that our satellite and other reach indicators point to a good house, rich village, good kids' education. It probably gives a better idea that they will do good in life and so forth. And

then they may be eligible for -- or the farmer in the family may become eligible for tractor loans. Good ideas about all these good land indicators and so forth may mean that they will be a better upsell customers as well. Gold loans are usually a flag, but if they are getting paid off early, that's probably an indicator of their resilience. A good house in the village and in general the village being in a good geographical area with good GDP, all of those can indicate early signals of a prime customer. So this is a thesis that actually brings up the Nostradamus concept that I'll talk about a little later. But before that, you may wonder, in rural India, are there enough digital footprints to really make all this happen? And in a traditional sense, in a traditional structured data sense, bank statements and credit bureaus and so on and so forth, perhaps not that much. But 90% of the data in the world is actually unstructured. And we are in a methodical mission to unlock value from them. An example, imagine a microfinance customer as part of their KYC process, our FLO has to go and get a picture of them with the house behind them.

And that KYC picture we can use with a multimodal LLM to understand the signals about whether this is a good house or not and give a score about their lifestyle index. We're using this today and we are taking those better scoring customers and upselling them into, or rather cross-selling them into micro LAP loans. This, by the way, we have invented and filed a patent on. That's not the only example. It is well known that microfinance customers who attend the monthly meetings are far less likely to default. So can we identify them? This is an actual picture of a microfinance meeting snapped by the FLO. We can take hundreds and thousands of these pictures, go through an advanced AI model based on deep learning architectures and understand from the existing KYC images which of them attended the meeting and who did not. And based on that, we can understand who are we going to upsell and who are we going to suppress? Not only is this applicable for images, but another kind of unstructured data is text conversations. One of the hardest problems in microfinance is identifying those customers who have the resiliency to continue paying a loan versus those who may flip on a bad economic situation. For that we are sending out psychometric questionnaires that don't have a clearly correct answer, but then we analyze them using GenAI and other technologies to give them a psychometric resilience score which we are in the process of evaluating as to how much difference they will make in loans. Now all of this, of course, requires a lot of data and you may ask is it worth it? So we came up with this concept of RODI, return on data investment. Simply put, how many rupees of credit losses are we suppressing for every rupee of data that we spend to acquire and process that data? It turns out that while the premium bureau sources are fairly good, has a pretty positive RODI, some of these other alternative sources have RODI that far exceed them. So this validates the path not just from a purely underwriting model improvement, but doing that in a sustainable way and that's why we are going after a host of other data sources in a structured ethical way to extract value from them.

Now we talked about portfolio intelligence, so project Nostradamus is going to be the first step in that where after onboarding, we keep taking a periodic peak at that customer and see whether they have improved their risk and resilience score in which case we can try to put them in the right buckets for cross-sells and upsells or whether we are getting questionable signals which lead us to then look up all of their available sources and understand whether there's a red flag and act on that by giving them to collection teams far before their DPD may actually turn positive. This is proactive management of portfolios and is only possible through data and models and looking at them very frequently. Now at the end of the day, we are in a B2C business and for that we have to talk to the customers in their own terms and give them value, let them understand our products. For that the industry has been trying to use chatbots. Generally, today's chatbot as they exist tend to be very non-conversational, they don't understand UL. And even if they do, a lot of them are not connected to the right agents underneath to actually bring out value and they sometimes ask you to do some very structured kind of data fill up like okay fill up this form or this particular format and so on and so forth and sometimes they are too menu driven, click here, click there rather than being something that is truly of a human-like nature. Instead we need conversational chatbots that actually converse to you like a mortgage officer would. We need something that is truly well equipped. You need an amortization schedule, here it is, download, take it and use it, bookmark it, come back. You need something where you can actually use this in a dynamic way. You want to talk to it in English, fine. Hindi, fine.

You want to ask cryptic questions, you want to ask longer questions, fine. It should dynamically adapt and finally it should be interactive. It should not be just a static menu. You should be able to see where you are and all that.

We have that as a demo today in the room next door along with four other demos of AI products that we have

built and we are in the process of revealing to the world. There will be more tomorrow where you are all invited. I ask upon Sonia, the Chief of our rural business to now come and talk about the wonderful things there.

Good evening, everyone, and welcome to the analyst meeting. Before we start, just a self introduction. So I have been in this organization for the last 16.5 years, managed diverse functions covering risk, collections, farm equipment finance business and in the last five years I have been managing the rural business finance which primarily comprises of microfinance as well as the joint liability group loans that we give. So just moving on after the digital as well as the AIML presentations, we will have a quick insight into the business realities and where tech and data is used and leveraged extensively to build the business. Microfinance as all of you are aware has evolved over the last 18-20 years in our country and moved from a business which was providing financial assistance to rural women in the underbanked areas. It has come a long way and evolved quite a lot. The construct of the business, the realities of the business have changed in these years. It has increased a lot of formal finance penetration in the rural areas. It has ensured that there is formal credit access to the rural population. It has also improved the digital financial penetration in these rural areas and helped women set up sustainable livelihoods as well. Now this business as all of you are aware is prone to a lot of event risk, whether it is natural events like COVID or natural calamity or even man-made events like what we saw in Andhra, in Assam, demonetization.

These are all man-made events. Irrespective of whether it is man-made or whether it is natural, we have seen that over the years this business has actually surpassed all these crises, has shown resilience in coming out of it, surviving and growing as well. That is the reality of this business.

Now what I am going to talk about is our journey in this business. We have been in this business for the last 15 years. Started off in 2008, completed almost 15 years in this business. Started off from the southern part of India.

What you can see over here, from the southern part of India we moved to the eastern part and then to west and then to north. So between 2008 to 2021 is how our journey covered almost the entire part of India, except northeast. So while we grew our business like this, our presence spread almost over to 16 states. We cater to around 350-odd districts, 1,80,000 villages. We have around 18,500 people who manage this business, including the field level officers and we also have 1,900-odd branches as we speak right now which cater to this business.

Now if you look at our growth that has happened in the last couple of years. In 2018, we were a Rs. 7,500 crores book. From there we have been systematically growing for almost a year and a half and we had the COVID scenario happening and for the next two years, there was a pause that we took on the business front just like what happened in the industry and post that the business has been growing in a very healthy way. Now as many of you would be aware, we stand among the top five financiers in the country of the JLG industry which is almost Rs. 4.5 lakh crores in size. We have a 1.3 crores captive customer base and 65 lakhs of active customer base today. Now if I look at the journey that we have built as a part of this last couple of years, our journey essentially on the tech and data side started off in 2018. So this was a time when this business typically was done as a very traditional brick-and-mortar model. This is a time when we envisaged that we need to have a sourcing app as well as a collection online sourcing as well as a collection app to be there so that we enable the field force to serve the customers better. So in 2018 is when we first started this industry. First we developed a sourcing app so that all applications get sourced digitally and then we also had an online receipting system wherein the mobile receipts would be issued by way of thermal printers to the customers. So that's something that we introduced way back in 2018. 2018 as many of you will be aware, we had a crisis in one of our prominent states in the east, post that we took a step back and we looked at what sort of credit risk guardrails and portfolio monitoring guardrail should be there to ensure that we don't suffer any sort of an eventuality like that in future. So the next three points that you see over here, the trigger-based early warning signal, this is something we developed consciously to track the leverage or the exposure of customers in each market to see whether the customers are borrowing beyond their means. And if it is happening, it gives us early signals of increasing heatedness in that market and hence we can take proactive actions to be away from that market. So this is the early warning signal that we developed.

The maker-checker workflow that we introduced, this was again industry first, this we introduced early part of 2019, is to have an independent assessment vertical which is not the business team who does the sourcing of

the cases and this would avoid any sort of bias that would come in as part of the business sourcing team's interest in achieving their targets. So this independent vertical which was separately assessing the business was almost rejecting 11 to 12 percentage of cases which get sourced which gave us an additional comfort that the quality of the business that we are doing is great. The last point is about validating that the customer who is onboarded and the bank account to which the funds gets digitally dispersed is one and the same. So we developed an algorithm which was checking the name of the customer from the penny credit reverse feed that we get from the bank, compare it to the name and the application and ensure that the bank account is the same. So this ensured that there are no dummy cases which are happening. So these three were done as part of the portfolio strengthening framework.

Between '20 to '22, as you are aware, we had COVID. So this was a step back when we took and we saw how do we build efficiencies and processes in the system. And this is when we understood that during the field verification, when the customer logs in the application and the application goes through the various internal rule checks and the credit rule checks, it's after a period that you are able to tell the customer that the customer loan gets rejected or approved and the throughput normally for this business remains around 50-odd percentage, 50- 55 percentage. So the remaining 45 percentage of customers where you actually enter all the data, there's a loss of time of the field level officer which happens. There's a loss of time wherein you are not able to give a decision to the customer. So we developed a system wherein we can give an instant principal credit approval at the field level by doing an instant bureau check. So this actually enabled improve the productivity and waste of time was avoided because immediately you could communicate the decision to the customer. We also saw that as part of our book during COVID that we had a large repeat customer base and before that we had a product which was towards the lag end of the cycle which is at the 24th month or the 21st month which is when a customer becomes eligible for a repeat loan. So we decided to take a step back and see which is the stage in the life cycle that the customer looks for a repeat loan and why not we have a product in the midterm. And we came out with a product which is at 15 months when the customer completes a cycle, but we ensured that we made the credit risk guardrail so strong that it can only go to a zero DPD customer, ever zero DPD customer, ODD customer. This pre-approved midterm renewal product was made available. We launched digital collections. This was post COVID. During COVID when our field machinery could not go to the field, for the first time, industry first, we had the QR code which was a dynamic QR code specifically linked to each loan agreement as part of the passbook so that the customer at the convenience of a home even if the field officer is not able to reach her was able to pay her installments and ensure that her loan remains in zero DPD. The last one was about the customer liveliness checks that we ensured. This was done by integrating with Digio by ensuring that the photo captured is of a live customer and it's not a photo captured of another photo or a dummy case which is getting captured. So these are all risk control measures that we put in place.

Moving on to the next period which is '22-'23. While I said that in 2018 we developed a sourcing app and a collection app, we took a call to unify both the apps so that the same officer who is on the field doing both business and collections will have the comfort of logging into one app and servicing the customer both on business as well as collections. We also had tailor-made pre-approved offers for customers. This was the customer pool. The entire customer pool will run through all the rule engines in the beginning before the field officer approaches the customer and a pool of almost 15 lakhs to 18 lakhs customer pool which qualifies all the credit checks, all the internal checks is provided to the field team so that they can do very focused bombardment of the cases and the repeat conversion improves. So this actually helped in improving the productivity quite a lot. We also implemented smart document management system by incorporating OCR and ML checks to ensure that the KYC checks happen, are validated well in time and immediately. We also built -- this was in 2022 when the regulation required us to do an income assessment of the customer, we built a system calculator, built trade activities behind in the system to have a calculation of the income happening at the backend when the input parameters are provided so that there is no chance of a manual bias or a manual error happening in the calculation. So this was implemented in March '22.

Moving on to '23-'24, as part of the document management system, e-agreement and e-signing, we moved into 100 percentage paperless journey. So today there is no paper that moves as part of this business sourcing till disbursement. We also integrated with multiple technology partners like HyperVerge, Karza to ensure that there

are KYC checks and validations happening immediately. We adopted geospatial technology. This is about looking at the data at a pin code level, a granular level of information to no pin code level, delinquencies pin code level, exposures pin code level, business expansion that happens, business growth that competitors has, every level of data at a pin code level. We check before we decide whether to be in a market, when to enter, when to exit a market. That's something that we adopted. And the last one, this was mentioned in the previous presentation about automated route maps. This was done by comparing the lat-long of the customer with the lat-long of the field agent when he goes out to collect and ensuring that in the shortest possible route he is able to cover maximum number of customers, which improved the efficiency.

Moving on to more current period, we are at the cusp of almost 25% digital collections from the time we started off doing this. There are four multiple options which have been given to customers who can pay through either the QR code or BBPS or an SMS link that they can pay. They can also pay through the PLANET app. We have introduced a DIY journey for these customers. This is for repeat customers who actually would want to get a new renewal of their loan. And when the field officer is not available, she can download PLANET app and log in her application, and the application takes the due course. So that's something that we have developed as part of our internal conversion. We developed a customer leverage tracker. This was extremely important because in the last one year, in fact, even by the end of almost 2023, we could see that the leverage in particular markets are increasing. Every market was showing a different this. And so we implemented a leverage tracker in terms of average customer leverage in each of these markets, what should be the threshold, where are we, where is our customer leverage heading towards? And then we risk categorize these customers into high, medium, low risk, and ensure that when we take an exposure, we take exposure on lower leverage customers. So this was implemented. And the last one is about building internal efficiencies by reducing manual work by ensuring that we do better credit decisioning. We utilize AI quite a lot. Now here it finishes off the journey that we had in building this business. Now if I clearly look at on the themes of grow, harvest, and protect, how did we scale up our business is what I mentioned over here. So the loan operating system that we have over here, the LOS on which the business happens is a homegrown system. And this homegrown system has the capability tool for login of almost 7 lakh applications a month. When you have 13,500 people working on the app at the same time, sourcing and collections, it manages it. So it's extremely customized for our requirements. We have an app-based digital sourcing journey, which I have explained already, 100% digital disbursement, and an instant credit assessment.

As part of data, I mentioned about looking at pin code level granular data to take credit calls, to take calls of geo- expansion. We also do integrations by way of managing the third-party service providers to ensure that the KYC validations are in place. What this has helped us achieve is what you see down there, which is about average monthly disbursements, which you can see over here. It's not working. So what you can see over here is that disbursement is at -- there's a 27 percentage increase in disbursement. Our monthly disbursement moved from Rs. 1,400 crores to almost Rs. 1,800 crores. The FLO productivity moved from Rs. 15 lakhs to Rs. 18 lakhs. That's another achievement that we could build in by improving the process and efficiencies. The process enhancement, the average log into disbursement time got cut down by almost 50%, 100% paperless journey, something that we could achieve and 12 lakh customers have already downloaded PLANET app and almost 2, 2.5 lakh customers use the PLANET app every month for services. Now on the right side, what you see is the way forward for these businesses. So on customer lifestyle index, some of the things which Debarag also explained. This is about understanding the affluence and the financial stability of the customer to understand what sort of income that she can generate and she can repay, and hence taking the right customer calls is what we are working on. AI-backed geo-expansion, beyond what we today look at bureau in terms of pin code and granular level data, looking at other tertiary data like the presence of a river, presence of a bridge, or the geography which is prone to flood or drought. These are other factors which we'll be integrating into before we undertake any geo-expansion. AI-based helpline, both for employees as well as customers for clarifying any doubts on the product or the policy or interest rates or the process is something that we are putting in place. Alternate data for these customers to enrich information about the customers so that we take the right customer calls is something that we are working on.

Moving on to harvesting existing customer base. As I mentioned, we have a very rich repeat customer base in our book. So what we have done is we have implemented straight through processing journeys for these customers so that those customers do not face any cumbersome journeys and have the easiest process to be

followed to avail a loan. We have also launched express journeys for Green Channel customers. This is a group which is fully eligible for a repeat loan, all of them qualifying all the credit parameters that we have defined for a repeat loan, which is an express journey that we have put in place. We're also working on a propensity model which will tell us what is the time in the life cycle of a customer where the customer would be looking out for funds and ensuring that we give them an offer at the right time? As part of data, I already explained that we have a pre- qualified customer pool given out to the field to improve field efficiency. We also use this entire customer database to ensure that we cross-sell the micro LAP product which we started off last year. The last one is about providing risk-based pricing. So what we have done is we have segmented customers' basis, the number of associations that they have and the leverage that they have, and then enabled differentiated risk-based pricing so that we give the best pricing to the lowest leverage and the best creditworthy and the lowest association customer. So an exclusive customer who has the lowest leverage outside gets the best rate that we can offer to.

So what has this resulted in? We have got almost a 40 percentage unique customer base, which is customers who have no other loans, only loans with us. We have a 67 percentage repeat base, which is both these are clear strengths of our portfolio. And we have 10 percentage of the micro LAP business, which happens today, happening through cross-selling from this existing base. Now on the right side, what you see is the way forward, which is predictive geo-analytics, which I explained. Understanding more about the customer by looking at the customer lifestyle patterns, the customer's behavior, the customer's tenure after taking a repeat loan and understanding more about the customer, assigning a score and developing customized product offerings for those customers is something that we are working on. We are also working on a loyalty program for rewarding customers based on their behavior with us.

Moving on to the last one, which is protecting the customer franchise. This is more about how do we build strong processes to ensure that our portfolio is resilient. So I explained about OCR and ML checks for customer identification and authentication, which happens. Early warning signals, as I said, early warning signals, not only on leverage, but also on-us and off-us, which is a customer paying to us, but not paying outside and the reverse.

Both are early warning signals, which comes into the business so that we can take conscious calls based on that.

Leveraging third-party service providers for KYC authentication is another one. Data pin code, data bureau analysis, I said. Risk detection using real-time data is something that we do as system-based triggers and flags that we have provided in the system, which goes as an input to the risk control unit on the field to do field intelligence and field checks to see if there are any other abnormalities that they see. And the last one is about analytics-based settlement model. This is about overdue customers who have already defaulted and moved into buckets. At what stage do you settle those cases? At what amount do you settle? So building automated foreclosures and settlement models is something that we have worked on. What this has resulted is a 99.5 percentage collection efficiency that we today have, a 95 percentage ODD, and a 97 percentage zero DPD book that we have today. And as I mentioned earlier, we have an independent appraisal vertical, which today is a testimony to the quality of the portfolio that we have been able to build. And the risk control team, which is an exclusive 700-member team, which is one risk manager per every three branches that we operate, who operate on the field, who give us the entire field intelligence for the business to take the right calls. On the right side that you see is the way forward, wherein we are building geopolitical models. So this is particularly about what are the potential of a market, what are the likely stress in a market, whether it is geopolitical, whether it is economical, and then take calls on geo-expansion in these markets. AI-enabled attendance marking is something Debarag explained in detail. Comparing the photograph of the woman in the group with the individual photograph in the loan application and seeing whether she has attended the customer meeting is something we are working on.

Repayment propensity model is something about customers, which are the customers who are today at zero DPD but are potentially likely to default because of increasing leverage from the industry. This is something that we do as part of our bureau checks. And the last one is about the psychometric assessment, which again Debarag explained, about getting more understanding of the customer so as to custom-made, tailor-made products can be offered to her. That finishes off the data and tech part of it and how we have built, leveraged data and tech, and we have built on processes, we build on the methods that we do business, credit risk guardrails to ensure that we have built a portfolio which is very differentiated and unique and resilient.

Now moving on, since we are having this meeting at the time when this industry is going through a tough situation, there has been a request from a couple of our key investors to have a business update as well. So in the next three to four slides, what you see is granular business updates, which normally would not be there in the Investor Deck, the last Investor Deck would not be there, one more deeper level of understanding of the data. So the first one that you see over here, this is about our monthly disbursements. So if you look at the monthly disbursements, our monthly disbursement trajectory had reached almost Rs. 1,800 crores, which is when we saw that there's a crisis in the industry. So we put collection at the forefront and said that if a branch qualifies collection metrics of 99 percentage, 98 percentage, we put different thresholds basis, the market where those branches are, and if the branch qualifies that and crosses that threshold, then the branch has the right to do repeat business. If it doesn't, then it doesn't have the right to do fresh business. So this sort of guardrails have been put in place and this ensured that the monthly business trajectory moved gradually from 1900 to almost like 1400. Now, on the second side, what you see over here is the association which is today of a lot of interest because once the number of associations that a customer can avail has got changed, customers have gone and borrowed from multiple lenders and that's something which has caused sort of stickiness in the industry. So when you look at it, we have 41 percentage of our book in LTFS plus zero, that is these are customers who do not have any other association and only loan with LTF. We have 27 percentage customers with LTF plus one association and 18 percentage customers with LTF plus two associations. So this is all that we fund. So our LTF norms continue to remain like this since the last one year. Now, MFIN norms still are at LTF plus three, our norms continue to be more conservative than the MFIN norms. We have eight percentage of book in LTF plus three and five percentage of book in LTF plus four. These are customers who after taking loan where I was the third financier has gone beyond that and taken loans. Now this is the book that will have to slowly run down and get to 10 to zero once the MFIN norms are implemented by all financiers. What we have also done as another measure is in the existing markets that we are operating since the leverage of customers have been increasing, we also wanted to take exposure on customers who have lower leverage. That's when we have explored Telangana and Andhra as two markets where the customer leverage is pretty much on the lower side. Western UP as well as Eastern Maharashtra, two markets where we have not been present. So we get more newer customers, more customers with a lesser leverage and that somehow we are trying to compensate for it. What we have also done because there is a collection stress on the field, we just didn't want our field team to be over stressed because of the difficulty or the time taken to collect.

We have augmented the field force by adding people, almost more than thousand odd people we have added to reduce the accounts per collector from 540 to 490 and we have also implemented a call center channel. This is an independent channel which calls up customers who default on the day of billing, on the day of payment and tell them, ask them why they have defaulted and take a PTP and pass it on to the field team. So that's an independent check which actually helps us know whether the field officer has reached the customer or not. What we have as part of our portfolio from what you see over there, 69% of our portfolio remains with zero association, zero external association or just one external association. So this is a big strength of our portfolio. What you see over here is more than or equal to four association as per MFIN norms. June ‘24 we had 7% of our portfolio in this which is Rs. 1,700 crores. By September it has come down to 1,371 which is a good sign of deleveraging happening in the market as a result of the MFIN norms coming in July where customers would not get any loans more than four loans. By October we are at 1250. So this is a number that we very closely track and these customers have been flagged off in the collection app of the field officer so that they prioritize these customers for collections. So since our collection cycle is in the first 10 days of the month, first 12 days of the month, they ensure that these customers are approached first for collections.

Now moving on to the norms by which we source. So as I said we fund as LTF plus 2 which is max LTF can be the third financier. So on our onboarding, so this is on monthly onboarding of cases, only LTF is 53 percentage, LTF plus 1 is 30 percentage, so that itself makes it 83 percentage and 17 percentage comes in through LTF plus 2. And as I said since we implemented these norms some time back, the LTF plus 3 has tendered to zero. For repeat loans is what we used to do it but even that we have cut off way back in April. So now we have the entire portfolio which is a new portfolio coming in is coming in these three categories of LTF plus 0, plus 1 and plus 2.

We have also added what I mentioned over here is we have also added a 500-member team for fresh sourcing.

So this ensures that these officers are only engaged in doing marketing activities in new areas which where we are not present and ensuring that we onboard new customers.

Now this is a slide I'm sure you would have seen in the last Investor PPT. We have just updated it with the October numbers. So our LTF plus 0 which is exclusive customer base is slowly increasing which is a good sign. LTF plus 1 is also gradually increasing. LTF plus 2 has remained constant. LTF plus 3 is coming down because we in April we have taken the call to cut off that for repeat customers. And LTF plus 4, 5, 6 and more than 6 that entire portfolio which is totalling to 5 percentage is coming down which clearly indicates deleveraging happening in the industry. Now if you look at the total portfolio 87% of our book is less than or equal to two external associations which is as per our norms and 95% of the book is compliant with MFIN norms.

Now moving on to the collection efficiency this again you would have seen in the Investor PPT. October numbers we have updated from 99.4 our 0 DPD collection efficiency has been at 99.34 in the last month. In 0 to 90 it has remained at 98.1. So the entire effort is first on completing collections and ensuring collection remains intact and the dip in collections is reduced to the most minimum. That's the entire effort and the focus on the field post which is when we do the business which is right now happening.

Moving on to the PAR so this was something that we often get asked from investors and analysts what our PAR 0 plus 30 plus 60 plus and 90 plus. So from Q2 FY24 what you can see over here Yes between 3 to 3.2. So from 3.2 our 1 plus which is you can say 0 plus has moved from 3.2 to 3.6 and all the other numbers have remained almost at the same level. The 0 plus which is 100 minus 99.34 which is my collection efficiency that is what gets slipped and is flowing in which is showing up in that slight increase in the 1 plus that you see which in those buckets which they land up in that is where the effort is to stabilize those accounts. And as we said in last quarter there have been multiple states where there has been flooding and a collection has been impacted because of that as well. So there are multiple reasons which are happening in the field. We are very consistently monitoring all of this to ensure that we manage our portfolio and ensure that our portfolio remains resilient. Now this is something we specifically added with this there has been a lot of talk about attrition in the industry and hence difficulty in managing collections. So while there was a stress in the industry the first step that we took was adding people, definitely because we had to reduce the load on their fellows the field level officers daily routine. Because otherwise he does not get time to complete his work. So that’s something that we did. We also took a step to improve to increase their salaries and incentives because when we did an industry benchmarking we saw that while we were in the top 3, 4 financiers our salary structure was pretty much on the lower side and hence we did a salary correction which actually helped in curbing attrition. We also added people and reduced the workload which actually reduced the attrition by 17 percentage between August to October. This coupled with a very strong supervisory layer we have a seven level hierarchy above the field level officer reaching from the field officer up till the business head. We have ensured that this layer of people that we have a supervisory layer is vintage people with five plus years of experience. Average vintage is around five and a half years. They are people who have seen the industry, been with us for a significantly long time, managed cycles in our company itself and know how to manage the portfolio. That’s one of our biggest strengths that helps us manage our portfolio so well. So if you look at the addition of people every market we have added depending on the dynamics of the market, the situation in the market, the events political events or climatic events which happen in the market. We have taken calls to add in different. Some of the high numbers that you see are in the newer states for example Maharashtra, UP and all what you see is the new geographies that we expanded where people got added.

Now moving on to the last slide. The newest product that we have introduced Micro LAP, micro loan against property. We started off in Tamil Nadu. Now we are present in Karnataka, Maharashtra as well as Gujarat as part of our expansion. We also intend to expand this to Telangana as well as Andhra Pradesh. Today we have a book of close to around Rs. 210-plus crores. We are operating around 50 plus branches with this product. So the inherent advantage that we get is that with the micro loans portfolio that we have with 18,000 people working on this with 13,500 people on the street catering to 180,000 villages. We get a lot of leads from them and it also helps us filter this base to see which are the potential customers for Micro LAP and thus do cross selling. So that's

something that we are focusing on so that we build a secured part of the book parallel to the book that we have on the micro finance side. That's it from the business update side and I would request Asheesh Goel to come on stage to present the Farmer Finance business. Thank you.

Thank you Sonia for the detailed presentation. Good evening. Thank you for taking the time out and coming here.

A little bit about myself. I've been now with the company for about five and a half years. Prior to that spent about 19 years with Citi. Primarily worked in the commercial asset backed finance space. And I'm here to basically take you through the use cases of digital, IT, analytics which were discussed in detail by Debarag and by Ramesh. But before I get there, let me just give you a brief synopsis of what the business is all about because that's going to put in perspective the use cases that we will discuss as we go along. So obviously we cater to the farmer segment across the board, small, medium, large farmers. We also service the agri traders and processors that are there.

We have a Rs. 15,000 crore odd book, about 11 lakh odd customers serviced so far. We work very closely with more than 2,500 dealers. As mentioned before, India is the largest tractor market in the world by units, so 9 lakh tractors odd get sold in India. So we work very closely with all the OEMs, your preferred financers with all of them.

The top five namely Mahindra, TAFE, Sonalika, John Deere and Escorts make up more than 90% of the total market that is there. And we have a very close association with them. Our product basket covers the entire range of things that are required by the farmer, right from the tractor to the implements to the upsell for the working capital requirement and warehouse receipt which was launched about a year and a half back. Again before I get there, one thing that had happened last year, year and a half, that we'd seen a negative demand as far as the tractor industry was concerned and that primarily came in because of paucity of rainfall, there was delayed rainfall, there was uneven rainfall, sporadic rainfall, water reservoir levels went down big time and all of that led to a negative sentiment around the industry. Now in the first half of the year, while there's been delayed rainfall but it's been in abundance about 108% over the long period average, water reservoir levels are at 87% of live capacity.

So it'll be prudent to see if whether or not those two things and more have really had a positive impact and one of the best ways to do that is to see as to how the second half of the year has started. So the first thing we look at is disbursements, the industry grew by 20%, our disbursements grew by 26%. As a matter of fact, last month was the highest ever disbursement done by us ever since inception of being in this particular business. The demand was not only in the new tractors but even in the upsell piece had an equal amount of growth at 26%. The other thing to note is that this growth did not come at the backdrop of adding fresh manpower. This came in by the productivity enhancements that we were able to get due to the interventions that we've had which we will discuss as we go along. The other thing of note is that the e-NACH penetration is up by 54%. It stands at about 65% right now. Now this is an important change because: a) it is change in the banking habits of the rural customer. And the second thing is it's also our ability to be able to get to his primary account, his operative account which leads to a higher clearance and e-NACH clearance today stands at about 78%. All of this has resulted in the first EMI net bounce at 2%. This is down from 6%. Now this augurs very well for the portfolio in the times to come because at 98% collection efficiency on the first EMI is a number that we were happy about. Typically, festive seasons are the month wherein due to personal consumption there is a dip in the collection efficiency. However, what we witnessed is 140 basis points betterment over the last year. The touch free collections are up by 15%. They currently stand at about 62%. Now this is also a big fillip on my opex because it doesn't require for me to go out there and collect and which obviously results in my on-due-date collections which stand at about 68%. All in all, all of this has ensured that I have been able to maintain the market leadership position in the Tractor Finance space.

Now moving on to the use cases, we are going to divide this into three. One is obviously going to be Protect, the second one is going to be Grow and the third one is going to be Harvest. So let's take the Protect piece first and this is going to be on our Tractor Finance piece. Like I said and a lot of you would know that we’ve been here for more than two decades doing this business. The association started when L&T had a tie-up with John Deere to manufacture tractors and that was the genesis of this business. Obviously, this has seen a change over a period of time. It was pen and paper that moved on to a scanned document, then it moved on to a tablet and then it

moved on to a mobile app which we have been using for the past couple of years which has made it completely digital and paperless. However, the challenges of this industry was something that we wanted to address and we continue to address. It is still a lot based on the declaration by the customer or the farmer. It obviously has to do with the fact that a lot of them are new to credit, hence there is a lack of data availability. Banking, like I said, is not something that we had visibility on and more importantly were two points that the assessment of the standard of living of the customer and also whether he was the person who was ultimately going to use the tractor or the fact that he understood the nuances of what it meant to own a tractor was dependent upon the field guy who would go and make that assessment. Now to be able to address these things, one of the things that we did was obviously and a lot of that has been discussed both by Ramesh, Debarag is utilisation of the KYC and the Bureau but couple of things that we really came in handy was our ability to be able to get his land records, his details of the land records. The kind of land it was, the kind of produce that was happening, the kind of prices that were prevalent for that particular produce, the past three years of performance. That along with geo data like the kind of electrification in that particular area, the road infrastructure available, all of that put together is what allowed us to build a rule engine. Now this is going to go into the next level and where Farm Cyclops which has been discussed is going to come into play by the end of this quarter. However, this allowed us for decile based decisioning. Now higher the decile, it was a straight through wherein no further manual intervention required and this enabled us to sanction in less than 24 hours. However, the ones which were higher decile and this is a video that you saw in Ramesh's presentation is wherein we were able to do a video PD in a real time basis. This was done again at two levels, one at the first level supervisory level and second at the underwriter level and with this what we were able to do was have a better understanding of the profile of the customer. Now what that means is that when you go and you have a look and the conversation is happening, you see a bike standing over there, you see the kind of house there is, you see whether there is a refrigerator in the house, whether there is a TV in the house. The entire standard of living comes out plus there is a communication that happens wherein we are able to understand through that discussion. And whether or not he is the person who actually is going to be taking it and whether he understands what it means and the EMI and the liability that he is taking on. And this is now evident in the number that I told you about which is the 6% first EMI bounce coming down to 2%.

The second thing is Grow and Warehouse Finance is something that we launched about a year and a half back.

Now let me just give you a brief overview of how this product works. A lot of you would know that. So the customer deposits the commodity at the warehouse, the warehouse issues a receipt and a quality report, basis that money gets disbursed to the customer, the customer repays, I give the release order to the warehouse and the warehouse then returns the commodity back to the customer. Now there are two, three things that we did which were different.

One we were only, we decided to only go with professionally managed warehouses which is about 40% of the total warehouses in the country. One of the reasons for that is professionally managed warehouses we were able to develop API integration because all the documents, there is no documentation that we take, all the data moves in through the APIs that we have had. And the second point is that these professionally managed warehouses give you indemnity against both quality and quantity and if there is a deviation from the warehouse receipt or the quality report, they are liable to make good the difference of the two. And the second is that we only chose 34 commodities out of the prevailing commodities to be able to do this business. Now in the first year, year and a half, we have been able to disburse Rs. 400-odd crores and because this is a short term product of six months, the AUM stands at about Rs. 150 crores. However, what I will take you through is, what was the reason that we wanted to develop this completely differently? And one of the issues was that this was a very document heavy business in the way it was being done. The second thing was obviously because of that the TAT was very high, sanction used to take anywhere from 7 to 20 days, disbursement would take 24 hours to 48 hours, the MTM process was completely manual and the release process was very cumbersome. Now to overcome these challenges, obviously we had a first of a kind digital workflow which was completely paperless. Now this was possible again because we were pulling data both on the personal side, on the financial side including GST, ITR and we had the ability to upload the profit loss and balance sheet. Entirely this would allow me to be able to run the BRE and to be able to make a decision. Now this decisioning we were able to do in less than two days, down from 7 to 20 days and we were able to disburse in 15 to 20 minutes as against the industry doing it in 24 hours to 48 hours. The other thing that we did was that the entire MTM process was automated wherein the pull would

happen at a periodic basis of the prevailing prices of the commodities which would then get run at the back end and MTMs would then go out to the customers as and when required. And the second thing is that the release process was seamless. We actually have a release process which is 24 by 7 and with that the customer, the release letter goes out to the warehouse and parallelly the same goes through mail to the customer for him to be able to take the release of the goods as and when he desires. Now this has ensured that we have been able to build up a sanction of Rs. 1,100 crores in the first year, year and a half of being in operation and the second thing is that MTM has only been 1% of the total disbursements that we have done. And we do not have a single customer who's overdue in the one and a half years of operation.

Moving on to the last one which is Harvest. Now I said we have about 11 lakh odd customers of which about 4.2 lakh odd customers are live. So it's very important for me to be able to cater to all their needs through the life cycles. Now we've been doing that and doing that well because if you remember I told you the upsell numbers also went up by 26%. However one of the things that we were currently doing was basically doing it through the field officer. So it was manpower dependent. A lot of the closure of the loan was happening basis that discussion.

So asset verification whether the customer still had the asset or not and how the entire structure of the loan would happen would be dependent upon that interaction. Now to be able to overcome that what we did was from a BAU process of the call centre calling up the customer, building up a lead and that lead being given to the field officer and the field officer then going and closing out. We changed the thing to the call centre calling up, sending him the link of the app, assisting the journey and the customer being able to do it themselves and the disbursal happening to the customers without the manpower being in between. Now it's a simple four-step multilingual journey which is there which the customer can do it themselves. This allows for the customer to blankly be able to get a disbursal and that too in two hours. That's it from my side. I would now as we move on from rural to urban request Sanjay to come on and take you through the urban piece. Thank you.

Good afternoon ladies and gentlemen. So just before we start just a short introduction on who I am. I've been with L&T Finance for the last two and a half years. Before that about 18 to 20 years I've spent between branch banking and lending, Kotak and HDFC. I think all of us realize professionally that and collectively what as an industry we have done in the last two and a half years in the Indian financial sector. We have leapfrogged and done more than what we did in the last 20 years. I'm also very happy to share with you that the way we have progressed in L&T Finance. We have not just been in the centre of the changes. We have also taken some of those disruptions and we have led a lot of those challenges. So with that the attempt here is to take you through the numbers obviously on the entire business. But more importantly what are the right to wins that we create and what are the challenges that we see in the industry. So urban finance essentially incorporates of three businesses mostly PL is a relatively newer business mostly between mortgages and two wheelers. And clearly that you see on the disbursement side Rs. 12.5 crores that's about Rs. 2,000 crores a month is equally divided between mortgages and two wheelers and about 20% on the PL. Because of the nature of the business and a mortgage being a long tenor you see it dominating on the book side and of the Rs. 42,000 crores you see about 52% coming from the mortgage business and the balance 50% equally divided between two wheeler and PL. What I have to give you a flavour of the kind of customer segments that we are into in all these three businesses. I've also given you the yields the tenors and the collection efficiencies to give you a sense of what kind of segments that we are in each of these three businesses.

The first business I will talk about is the Two Wheeler business. So this is a -- I think this is one of the most intense businesses in our country. We do about at a country level we do about 1.7 to 1.8 crore two wheelers in a year, amongst the top two in the world. So while the industry has grown by about 10% and the industry has recovered post COVID. We have grown over faster than the industry at about 24%. So there are three large pillars that we have the entire two wheeler proposition is riding on. One as we all realize right now the customer acquisition still happens at the dealership which means that the dealer partner is the most critical person. So we have a very strong value partnership program and I will talk about that and how we have leveraged technology around it. We

have not just restricted ourselves to channel partners we are also preferential partners with certain premium OEMs along with two to three electric vehicle OEMs. We also have a very strong understanding of digital ecosystem and because of the intensity or the scale in which this business is, we have been able to create a right to win on creating a digital almost a paperless journey. Going forward we see that the industry will continue to grow at about 10% to 11%. LTF right to win. I think we will roughly be about 20%. Three key levers here that I would want to talk about and I'll share with you what we are doing on the valued partner program. So this is not just a partner program which defines the remuneration that the dealer gets but also beyond remuneration what is in it for the dealer. And the entire portfolio quality depends upon or decides what kind of partnership or what depth of partnership we will go with the dealer. Two because this is completely digital and there is no paperwork required there is no pressure on the opex when we are going to newer markets or there is no, as long as collections is taken care of there is no acquisition opex that we are getting into. And finally two words that you have heard a lot about through the previous presentations is Premium and Cyclops and these are the two cornerstones on all the businesses and I'll explain why.

So a tough slide but I think Debarag and Ramesh they explained it so my job is much easier. I'll just explain the simple part of it. The tech is already taken care by them. So this is on the left hand side this is where the sales guy comes in. Everything on his app or tab there's no physical interaction of documentation with the customer hence no concept of actually a branch required everything happens at a dealership. This is where the entire decision making comes in. So whether we have two metrics four metrics the number of variables that we get and I'll just explain a little about this and this is finally the disbursement to the dealer. So the moment everything is approved the immediate disbursement happens to the dealer. And again no paperwork and I'm just explaining because this creates a very low opex business model. Now this is where the key thing is that, do we have enough confidence on the customer to be giving a decision in less than two minutes and that's what the entire decision making does. So what we have in here is either to a good customer where we have multiple inputs, we will either give him a better LTV or a stronger offer. Where we are not confident, we will either reject the customer or the LTVs will be much lower. What hence it leads to is a completely digital business no manual underwriting as you all see and most critically the entire decision making is automated. On the valued partner program this is an industry first and we realize that there's a strong reinforcement that needs to happen at the dealer partner because what beyond the payout that we give to the dealer. So in phase one what you see this is the value partner program currently managing about 9,000 active dealers out of which value partner are about 2,000 so we realize that this kind of scale is very critical to be going through technology. Right now what we have built it which is phase one it takes care of the business it takes care of the portfolio health because that is what the portfolio health decides what eventually how deep the relationship will be with the dealer and finally what the dealer gets along with the trade advance. So in this business trade advance is a very critical factor of doing business. In phase two we realize that this is a sales efficiency intensive model and we realize that because there are humans in the entire sales process there will be inefficiencies and there will be gaps and the phase two will create a two-way communication with the dealer. So the reliance on the sales channel will be much lesser because of this app.

How it will look like or this is actually in place, so this is how the achievement is which means that what he gets this is how the approval rates are on the right hand side top corner, left hand side bottom you see the portfolio health and portfolio health decides what kind of how far in the relationship we go with the dealer and finally on the trade advance the right hand bottom corner that you see.

Two-wheeler is not just a product for us. Two-wheeler is also an acquisition or a cross-sell or a data mining because we have a very strong information on the customer very good understanding. We also realize that two- wheeler is normally the first asset product that a customer takes. And if we do a good job in terms of taking care of his requirements and do the credit properly, we think that this is a strong digital native which will help us build on the journey with the customer rather than being a standalone product. And that's where the entire credit mechanism comes in. So about 7.3 million customers that we have on the two-wheeler side. We don't just look at the bureau and simple guardrails there are multiple portfolio actions basis which while we are talking there are about 8.5 lakh customers on the two-wheeler approved for a personal loan. Just to give you a sense of what we have done over the last and this is about a three-year-old product. We have been able to disburse close to about

Rs. 9,300 crores so far an active book of about Rs. 4,000 crores. We roughly every month do about 1.2% to 1.1% of the entire Rs. 8.5 lakhs that you see for personal consumption. Prime which we have already spoken about I am reinforcing, prime is not just bureau score greater than 730 and above. A prime customer is a customer who has at least two or three of these metrics. So it could be an EV customer with an account aggregator or having a certain bureau score with a certain income level and a certain LTV so it's a mix and match of at least two to three variables. And clearly you see by design how the prime share in two-wheelers has grown, clearly from 39% to 64% almost doubled. We continue to see this growing very aggressively and Debra and Ramesh what they spoke about Cyclops. So two-wheeler is the first used case for Cyclops and you see the upside going essentially because of this. While we are talking about 50% of the entire business in two-wheelers happens on Cyclops. We intend to take 100% in the next two months. Already prime share is up by about 10% LTVs and ATS are up by about 4% to 5% and GNS is down by about 55% to 60%. Net-net what we have ended H1 at about book growth of about 33%, this will continue to grow I've already shared the numbers with you. With Cyclops at 56% and prime at 64%, with Cyclops now going to 100% you can imagine the kind of prime that we are going to get, along with a drop on the NNS.

So the second business that I would want to take you through is the personal loan business clearly left top corner you see the industry has gone through a challenge. For LTF most of the PL business has come from the internal data mining that we did on the two-wheeler customer. So close to about Rs. 7,000 crores book much smaller than the rest of the book. Three key things that we or three key pillars that we worked on, one was that we had very strong understanding of a digital journey on the two-wheeler side and we replicated it on PL. Two, we worked very closely with analytics and credit on how do we build offers for the customers so what we already spoke about 8.5 lakh customer offers and finally we went out into the open market and we said some of these digital journeys can be integrated with open partners with this kind of credit infrastructure. So that we create a product where we are able to without diluting the guardrails we are able to sanction at less than two minutes. We see with the consumption, the way it is happening in India, we see that this product will continue to grow at about 14% to 15%.

LTF finance will follow that growth rate. However, the levers that we are working on we will continue to upsell on existing customers right now we do on two-wheelers we will get in other products as well. Second the physical channels that we are targeting this is 100% salaried, prime customers, category employers this will continue to grow. And remember the journeys that we are getting to the market even with these channel partners who work otherwise with large banks has been able to get us the right to win.

Finally we will grow with large mega partners and I'll take some time to explain what a mega partner is. So mega partners are partners who have customer base of 10 million and above. So that's the first criteria of a mega part what does mega partner get to the table superior UX a recurring customer base but along with this it also gets a very strong customer demographic and a behaviour understanding. What we get to the table, we understand the product very well, we understand the guardrails very well and between the partner and us the intention to create a journey which is integrated rather than broken. We feel that with this kind of a model we will be able to create a very seamless customer experience resulting in higher conversion on the salaried and the prime base. Sharing with you a case study with a mega partner which fits into all this where we went live about four months back. And that gives us the confidence to go on while you see the scale-up that has happened in about four to five months.

I have just indexed it. The kind of credit cost we are at a zero NNS in the last four to five months. On a premiumization, this is 73% salary based and about 80% premium. So that gives us the confidence that as and when we go with more partners, we will see more upside on this prime base. So thankfully I don't have to talk about it, on the left is our existing, on the right is the Cyclops. It's already been talked about by everybody, so I'll not touch upon this but Cyclops will be launched in personal loan business by end of the year. The way forward on this business Cyclops going live, we will be integrating with three new partners other than the one mega partner that we are integrated with. There'll be a launch that we'll be doing tomorrow with one of these large mega partners and finally the channel and geo expansion along with cross-sell opportunities because these are completely digital products there is no unnecessary stress on the opex, we'll be able to scale up very quickly. Eventually you see 77% on existing customer that's the personal loan H1 and premiumization and salaried continue to go up. NNS reduced by about 55%.

The final product that I would want to talk about is mortgages. We realize that industry still does mortgages like a very traditional product and it is brick and mortar because eventually you're buying a brick and mortar property.

What however we have been able to do is that because of our experience of digital because of understanding today we are one of those very few who have been able to create a complete end-to-end digital journey. So what that does is while we also work with external partners like a lot of industry does, the kind of journeys that we are able to give to the customer and Kavita who's the chief marketing officer she'll talk about it in our subsequent slides. We said we don't want to be another home loan provider in the market. So what is it in it for the customer.

And we identified the genesis Kavita will talk about but I'll just talking about the output. We said there's a home decor which means most of the customers, the result paying the down payment and the everything and they are left with very little to do up the house in terms of the interior. So that's where the home decor comes in. This is for all our home loan customers. The second we said, that the way SKY RM the way a branch banking or a banking operates. Home loans are typically 15 year to 20-year end loans. And during this tenor while the customer doesn't require too much of interaction there are points where the customer requires and those are the points where he gets upset and that's where the balance transfer happens to another bank because nobody takes care of that customer. So we have built a central RM which is right now at Bombay. So 60% of our entire customer base that we acquire has a SKY RM dedicated to take care of any customer requirement the way a relationship manager does in the branches. Current offtake of home decor on our eligible set of customers is about 15%. We also feel that with these two right to wins other than the digital there's an opportunity hence to mine our existing customers.

Right now, we were mining two wheeler customers for personal loans. We believe that there's an opportunity to mine our entire base. We have identified about 2 lakh customers in our existing base who have Rs. 62,000 crore of home loan opportunity. The offering to these customers is not just a better price. It is better price plus it is a home decor which takes care of their requirement plus it is a SKY RM plus it is a completely digital process. With this we think that we will be able to create a very strong proposition in an otherwise undifferentiated home loan market. So on share metrics this is what we ended H1 at book growth of about 42% in mortgages, completely prime. Credit cost is sub 0.2% in this.

While I come to the end of business presentation, there is no presentation right now complete unless you talk about collections. Clearly if you see most of our businesses, we have indexed it against industry do exceptionally well and it's not just the output. If you look at on the right-hand side, I've just put together how we are allocating, how the entire delinquency management is happening and analytics and AI is a very integral part of how we allocate because the intensity of how many calls a customer needs to be reached out to, which customer requires a face-to-face versus a telesales. So right now 76% of the customers due, are taken care through an AI model or self-cure. Between 18 and 6 that's about 24%-24.1% customers are ever called by a physical either a bot or a voice for a due date and that too out of this 18% is taken care by a contact center. So again no physical people going and no friction points. So roughly about 6% that you see on the right-hand corner is what is actually through a face-to-face interaction with the customer and we feel that it is very critical because these are the friction points in the market. We have an indigenous app, a BRAKE App which takes care of the entire hands-off between us and the customer. Just to give you a snapshot of how the BRAKE App looks like. A lot of it is outsourced to agencies and we have about 4,000 to 5,000 collection executives who are hired not by us but by the agency.

Every executive out of these four and a half thousand there's a KYC done and there's a separate login ID created on this BRAKE App and every customer which is allocated and called will have to be first allocated on the BRAKE App and there is no other way of doing collections outside the BRAKE App. What it also does is it gives the individual a complete understanding on what is the customer that he's going after and the supervisor for us at a central level. We have complete control on who is meeting the customer, what point of time. Eventually the money is collected, the last mile, the slip-ups, the friction points, everything is taken care of. Everything that we do in collections or discussions and reviews happen completely on the BRAKE App. So with that I come to the end of my Urban Finance presentation and I would call upon Abhishek Sharma. Abhishek, who is the Chief Executive of the SME business to take you through the business fundamentals. Thank you so much.

Thank you Sanjay. Good evening to all. So let me start with a brief intro of mine. Graduated in the year 2000, did a stint of around six odd years with Ministry of Defence, then two years in Jamshedpur doing my MBA and last 15 and a half years serving with L&T Finance and what I'm going to introduce now and talk about is the SME business of L&T Finance.

So this business was conceptualized and envisaged as part of post-COVID Lakshya 26 target that we rolled out.

And post around one odd years of research and FGDs, what we came up with, was that we started our pilot in two locations. The focus of the pilot and why we wanted to approach it like this, is that SME is a very, very diverse universe and we wanted to exactly understand what are the customer needs which we need to cater. Second aspect was that when we approached this market there were already more than 40 players entrenched in this market. And hence we needed a value proposition which is slightly different which is going to attract customers towards us. And third is one of the more important person or one of the most important person is the channel because this business is generally sourced through channel and hence we needed a value proposition for them as well. So with those aspects in mind and a two around good nine months of pilot, we came up with what exactly we are going to do and with that in phase one we had around 17 odd locations and we did a book of around Rs. 1,500 odd crores which was built as part of phase one. Subsequently in phase two another 54 locations were added and our manpower was around 300 odd and a book size of Rs. 2,800 odd crores and finally the current phase where we are around 110 locations plus we are present with a manpower of close to 400 odd with a book size of around 5200 odd. Now one of the things that we understood and that on the left side that the sequence of product launches. So when we launched initially it was only a plain vanilla term loan. However we understood that flexibility of repayment is something which is valued by the customers who are prime and super prime and hence we came up with drop line OD and hybrid OD product. Just to explain what this means is that within the given principle which keeps declining month on month a person has the flexibility of paying more than his EMI but can withdraw whenever the need arises. So that's the construct under which this product approaches. By March-24 we also had a decent customer base who had serviced their loans over a period of time. Hence as a natural outcome Top-up was launched in August. SuFin is a L&T's own industrial B2B marketplace. So we had a tie up with SuFin and through that in August-24 and finally what we expect is in next year in the next month rather in December, we will be launching our supply chain finance business, which Ramesh also talked about.

Coming to how we approach this. What we understood out of the market out, these are the five big pillars under which the whole business stands. So first is that it's a primarily digitally native business. What it means that there is no paper there is no physical signature as far as customer is involved. What has been achieved through is integration through 20-odd APIs. Also there are multiple portals for all the stakeholders who are involved in any particular transactions and hence from a customer service viewpoint it becomes seamless. So that's the pillar one. Second pillar around which this business stands is that from day one we have approached this through a risk-based pricing. So around 200 basis points difference exists between near prime and super prime and for different customer segments there are different customer journeys as well. Third is around location selection.

Today as we speak from 110 odd locations we service around 18% of the overall market and wherever we are present we are having around 5% plus market share. Now how we have selected these locations is on three dimensions. First is the underlying market whether the size is good enough or not. Second whether it has been growing and last but not the least whether the delinquencies are controlled in those underlying markets or not. So under those three dimensions we selected the locations where we are present at this point of time. The complete portfolio goes for every month as scrubbed and hence we have very strong EWS measures. The output of that is our first time bounce is range bound for last 18 odd months in the range of around 2% odd. Last, but not the least that, while we do have 150 odd BRE rules we strongly believe that on-field presence of credit gives us strong inputs about how to approach this business and hence in 75 locations odd plus we have credit managers manning those locations as well to provide those inputs. So hence what we believe is that generally we do have a plug and play model available for expansion as we move forward.

Now coming to the resilience part of it if we index ourselves against industry, our standard book which is a 0 DPD book is around 106, vis-à-vis, if we consider industry as 100. As a natural outcome of that in the buckets our book is lower than whatever is there in the industry. How we have managed to do this is that if we see on a ticket size cut, the major delinquency comes from 5 lakhs to 10 lakhs bracket and we have completely stayed away from it.

So our ATS is around 25 odd lakhs which happens to be the sweet spot of the industry as well and that has allowed us a leverage here. Second is that there is complete absence of subprime in the portfolio. We do not do any case which is subprime. Our offering starts from near prime and above. Even on CMR if we see our distribution it is primarily between low risk and medium risk and hence we have been able to achieve a range bound collection efficiency over a long time horizon between 99.7 to 99.6. So that is how the resilience of the portfolio has been.

Now when it gives us reasons to believe that one we have a customer journey which is amenable to changes that customer requires and second is that we do have reasons to believe that we have a resilient portfolio.

So what next now and going forward what we intend to do is one, Project Cyclops naturally. What it allows us is that today we take these four as our primary data points which is his banking, bureau, GST and application data.

We intend to add the GeoIQ which gives us an insight on where exactly this business is located and what is the business intensity, the signals around his payments and other digital footprints that he might be having and which will fundamentally allow us to do a better improved customer segmentation while strengthening scorecard and real-time decisioning. Another aspect that will be adjacency that we will be launching is around supply chain finance. We all know this business so hence I will not get into explaining the business but the focus of our implementation of supply chain finance is that we have to take opex out of this particular model to the extent possible and hence we have invested in a state-of-the-art platform which will allow all the ecosystem players to be on the same platform and hence reduce the cost of communication between each partners and reduce the cost of credit as well. How we intend launching is that, initially it will be first with dealer finance then followed by vendor and then subsequently a bill discounting model will be launched. Next aspect is around direct channel scaling. Currently 94% of our business is through channel. We would like to increase the 6% which is a direct contribution through a combination of cross-sell, tie-up with other fintech partners. Second is, continue our effort as far as ease of customer use is concerned, PLANET app 17% of our OD customers every month transact through PLANET app and take their withdrawals. So that's something that we will continue to build upon. Last but not the least is that we would like to get into the adjacency of secured BL both through form of an insurance guarantee scheme which is CGTMSE as well as through other non-property collaterals and machinery loans and all, thereby giving us a reduced cost of credit through a cycle-resistant portfolio.

Coming to the last bit of it, our NPS stands strong at 62. We have engaged customer base and around 70% of our customers do use PLANET app regularly and we continue to work on three important value propositions that we are trying to build in. One is that continue to have a fully digitized process, second is quick loan disbursal and third allow flexibility of repayments to the customers within the overall ambit of a declining principle. It is always better to hear from the customer what they think about us and let me conclude with a customer testimonial around it. Thank you so much ladies and gentlemen. With this I hand over the stage to Mr. Raju Dodti, our Chief Operating Officer. Thank you

Good evening, all. Just when Abhishek was talking, I was having a word with my CFO, Mr. Sachinn Joshi with whom you all must have interacted. I asked him what is that one sentence or one thing people from this room should walk out with and he mentioned something to me. I will come back to that. Exactly eight or nine months ago in one of the investors meet, we were visiting certain investors, Sudipta was there, me and Sachinn and that was the first time Sudipta uttered the word Cyclops in front of us. To be very honest with you, coming from the world of legal, I was a general counsel in this company at one point in time. For me to get understanding and affinity with that word was bit challenging. But in this last eight months, the organization has gone at the cost of sounding little pompous, I must say, through a transformation where every single employee in the organization today is virtually eating, drinking and breathing digital and technology and that is courtesy three individuals. I must

mention them, Sudipta, Ramesh Aithal and Dr. Debarag. A big round of applause for these three individuals for bringing the organization to this stage where we could get a confidence to dedicatedly have an Investor Digital Day for all of you to let you know what exactly we are doing in our company and how it is impacting and transforming the company. I must also say thank you to the customer whose testimonial we just saw a while ago.

His name is Harsh, which means happiness and joy, which is the exact feeling I am experiencing at this point in time as I am standing on the podium in front of you. Three areas where technology and digital initiatives of ours have impacted the most are operations, customer service and credit. Our business executives took you through all the businesses in granular details. I will take you through in briefly what are the initiatives which we have taken in these three areas and what exactly has happened because of that.

Briefly, I joined the company around 10 years ago as Group General Counsel, led three businesses as Chief Executive Infrastructure, Real Estate Financing and SME and have now assumed the role as COO recently. When it comes to the operations, we have been successful in setting up a tech enabled 24x7 operation desk and the whole idea of the operation is just not to handle transactions but get them at the volume which would be seen multiplying as we pass many years. We were almost one third exactly three to four years ago. We have been able to achieve the volume of both the transactions overall as well as the disbursement which we are doing. We are probably disbursing almost Rs. 50,000 to Rs. 60,000 crores of disbursement in a year only in retail segment. All this has happened whilst we tracking two key parameters. One is the operation cost per transaction and the second is the cost per disbursement. In the bottom part of the screen, you would see both these numbers. I should not utter them once again but this has happened because of few initiatives which we have taken. When it comes to the disbursement, we have been in a position to manage bulk of our disbursement with the STP mode. So these are totally touch free. Right from the time a customer approaches our desk for a loan, till the time he gets the first rupee of disbursement into his account, almost 50%, the precise number is 46% of the disbursement happen through STP mode. Our endeavour is to take this number to 100% but as you understood from last conversations, we are operating eight business lines in the hinterland of the country in both rural and urban across 2000 locations. So it does take a time to take it to the 100%, but that is what our endeavour would be. Moving on, when it comes to the loan closures, earlier the loan closures would take customer to approach the branch or at least call the customer call centre on a phone and spend a bit of a time over there. Today we are in a position, we probably would be one of the companies who can definitely take a credit of managing almost 100% of the transactions of loan closure happening on the STP mode. We have out of around 30 processes to 35 processes which go right from the application of the loan to the closure, we have almost automated 25 processes out of that.

All these three key initiatives has led us to basically get the benefit which we are seeing on the screen in terms of the increased volume of transaction and the falling cost per transactions.

When it comes to the customer service, we launched the net promoter score initiative in two broad areas. Firstly, we thought why not start given the fact that we have almost taken most of our initiatives on the digital platform.

Let's start with the onboarding of customer. We began from there, I'm very happy that we have reached a score of 55 odd percentage. This is one data on which basically the opinions are pretty diversified and therefore I went to that one source where we get the maximum amount of inputs nowadays and that is ChatGPT. I asked what is the good score for a digitally enabled NBFC in India to have when it comes to the NPS and I got a score between 40 to 50, I was very happy, but having said that happiness in customer service will not take there, we have internally set the threshold of at least 70% when it comes to the NPS of onboarding. Having said that, we also moved and took initiated another NPS for customer service and in that we have got a very massive score of 80% plus and we are satisfied with that. Will we remain absolutely satisfied and therefore become little relaxed. The answer is no. We will keep on investing in the customer service and therefore we have enabled multi-channel customer portal to interact with us. Everything is there, be it a Chatbot, be it a web enabled services, be it our PLANET app about which Ramesh took us through in the detailed manner, but most importantly we understand that when it comes to the customer, artificial intelligence is the one on the basis of which we will pivot. But we will never forget the emotional intelligence part of it and therefore we have kept almost 150 customer service executives in those branches of ours so that customer has an option. She can definitely go and meet our customer executive to get that service as and when needed. When it comes to the regulatory complaints as well as the

customer complaints, we have seen all these initiatives which we saw on the first slide and what is there in front of us is leading to per thousand complaints falling massively down. This is one of the big barometers we will be completing and we will be keeping a watch on to ensure that all our initiatives in this area are on the right track.

When it comes to the last slide of mine about the credit, everyone, every single gentleman and Sonia when they came they spoke about credit as Sudipta said credit is our middle name. I think Sudipta we can say credit is our middle name, our first name and our surname. At the end of the day that customer who gives you the best business is the one who will eventually lead you to the situation of a lower credit cost and therefore from the old age credit underwriting model which was manual in nature to begin with and which was based on bureau score and some amount of information being looked at, we have moved to the credit engine wherein we have got all this information ahead of us. So we have a bureau data, we do have the account aggregations data and we have that score coming from the geolocations which gives us the various kinds of indexes which help us to do the segmentation of the customer. This has led to many benefits for us. One definitely the quicker turnaround time, second is the improved credit cost for us and last but not the least it is allowing us to offer best offers, in the best interest of the customers and with which we are operating this franchise. In nutshell, these are some of the few initiatives which we will be focusing on in the next 12 odd months. The big one in this is our 24 by 7 tech enabled operation center is already operational and in next 3 months we will be completely optimizing it and it is based in Mahape. We would be moving mostly the customer service is outsourced to the call centers. We two were in the same league.

We are moving external call centers to the captive internal L&T finance employees handling it because we want to believe that the customer service would be the one that should remain at the heart of everything that we will be doing in L&T finance. Without any ado I would not have any shame therefore to subscribe that one of the profound quotes on the customer service by none other than Jeff Bezos which is on the screen. Thank you so much. I will now request Kavitha our CMO to take the podium

Thank you Raju and good evening everyone. I'm Kavita Jagtiani. Well by profession I'm a marketer. I'm an marketer and my passion lies in building winning brands. At L&T Finance, I have spent about a little over a year, but I have about 25 years of steering the marketing function in leading BFSI and consumer products companies.

Well to start the presentation I must say this crowd has a lot of action, but as we speak there is some action that is happening at the India-Australia cricket match right now in Perth. Now we cannot go there, but can we bring some of that action here? Let's see. Well as you just saw this is 3D animation and with the advent of 3D CGI, augmented reality, GenAI, it has all added a whole new dimension to marketing and that is what I'm going to cover in the course of my presentation, while I will bring our Brand Ambassador back into the presentation later. At present, I'm going to cover the home loan campaign. You saw some of these visuals during the home loan TVCs that were aired at the start of the session. Now I'm going to talk about what led to the creation of this campaign as well as how did we execute it with the use of digitization. Well before the campaign we had very low level of awareness and therefore the objective before us was to increase the awareness through repositioning our offering.

Now how we went about doing that is we took up a research to know what are the kind of pain points that the consumers are facing and they remain unaddressed by the category leaders. That is why we decided to fulfill this gap and disrupt the entire home loan category with the introduction of “L&T Finance Complete Home Loan” which came with these benefits which as you saw Sanjay elaborated in his presentation. Well what was important for us was to drive this category message across all stages of the consumer journey starting from awareness.

When it came to awareness we took up both digital and conventional form of advertising. In digital, we were present on connected TV. So connected TV we were there on JioCinema which was airing the IPL. As you know connected TV is a digital serving medium as compared to linear TV. So it was important for us to understand what were the kind of impressions that we were getting for each of the creatives that you saw especially in the key markets where the match was being played. Now how is that possible? This was possible with a tool that's called a seismic tool which tracks every impression during the IPL. So there's a seismic pixel tracker which gets embedded into the digital files and therefore every time the impression is served we know that. It also measures the kind of unique TV sets where the impression is served as well the entire duration for which the creative has

been played. So of course we were able to achieve about 215 million views because of this tracking. We also took up outdoor advertising which is essentially billboards. Now here too it was important for us to track what is the kind of reach we are getting. So the tool that we put to use was mobile propensity which helps us to track the outdoor viewership. What this does is that it looks at data from different sources. So, for example, it looks at mobility data which is essentially all the GPS logs that have been logged in through the mobile apps in that area along with the census data and the profile data to estimate what is the reach that we are getting for every billboard.

So this therefore helped us in precision delivery of our media. Moving on we also took up for discovery and consideration Google search and influencers and here we utilized the Google analytics to understand the kind of source keywords and the combination of keywords that would give us high volumes as well as the kind of traction that we are getting from the contextual influencers. Also I'm glad to share that organically for the keyword complete home loan, L&T finance home loans ranks first on the page one of the search. So therefore conversion.

Additionally through a lot of performance media and retargeting we were able to drive higher conversions and personalization. Now basis every intent so for example if a consumer is logged in to magic bricks to search for property or has applied for a home loan. Now that is the time we personalize messages and that happens through AI and ML based dynamic ads which are triggered and of course we generated considerable amount of leads.

Now this entire exercise helped us to drive higher levels of awareness so our top of mind awareness improved from 1% to 14%. So the model I spoke about was all about driving marketing at an aggregate level. However, this is the model which now helps us to drive marketing at a single customer level. The model what it does is that it looks at data from different sources. It also adds on the propensity modelling and segmentation which gets integrated into the unified customer data, integrated with that are the MarTech tools which enables a single view of the consumer. Further through AI we are able to segment and drive higher ROIs through hyper personalization.

Now let me explain this to you with an example. So what we've done is followed this hyper personalization for our two-wheeler customer onboarding journey. So right from the start when the lead is generated and up to the disbursement, we are able to track the consumer as well as understand what are the kind of interventions that we can do. So because of the single view or the 360 degree view that I spoke about we're able to understand at what points of time we can do an outbound call through the call center or also try to do a whatsapp or an email message depending on what stage the consumer is at as well as hyper personalize the communication. So, for example, if let's say a consumer has served a lead and we want to nudge him to move to the next stage here's what we do as a message. So as you saw we have got Bumrah back and if let's say your Anurag and your loan is sanctioned here's the message that you will get. So this is the kind of hyper personalization we're able to enable and of course that helps us to drive higher login to disbursement ratio. Moving on there is great use case of AI both for image and content creation and we've used that for a recent festive campaign for Diwali which was an AI enabled program called Sapno Wali Diwali with L&T Finance where we set up an AI led microsite, had consumers visit the microsite, select your dream, upload your picture and have an AI generated image which got us a very huge response from people and as you can see these are all the kind of images that got created all thanks to AI and GenAI and we were able to improve the level of engagement to as high as 23% and get 16 million views on social media. If you want to know more about what GenAI can do in marketing do visit the marketing stall which is at in the next room. Moving on just beyond image there's also a great use case of AI in video creation and we did just that. So the event which is coming up tomorrow our marquee Raise Event. We have an entire AI generated video that talks to you about what the event is all about and this gives you a sneak preview of what is in store for you tomorrow. Happy viewing and we look forward to seeing you at RAISE tomorrow. Thank you and now I request Sudipta and Sachinn sir to come on stage to address Q&A

Thank you so much Kavita. Thank you so much everyone for the granular perspective on LTF. We now move on to the Q&A session. I now invite questions from the audience please. We request the audience to restrict their questions to one for maximum participation. We have volunteers who will assist you with mics on each side of the hall. We request you to provide your name and organization before you ask the question please. So Sudipta sir and Sachinn sir kindly please come on stage for the Q&A session. While they come in let me tell you that the

lanyard you're wearing is your pass for tomorrow for Raise’24. So please do attend RAISE 2024 with the lanyard that you're wearing tomorrow, that’s your pass for tomorrow. Thank you,

Long presentation, 162 slides.

My name is Digant Haria from GreenEdge Wealth. So Sudipta great presentations and these two examples specially that undiscovered prime and the second example of the AI application which was if the customer is good you have lower down payments, higher LTVs and different interest rates. Those were great examples, but so my question - it's a question and compliment both by doing so much of analytics, so much of capabilities will we be getting enough customers to grow at 25% because a lot of your competitors would actually be lending without doing any of these things. So that's my question.

Yes so let me take that. See I think India is a large country. There are lots of customers out there so I don't think we'll have a dearth of customers, but I'll tell you where we'll get enough customers. Right now we let go of a lot of customers because we are not sure of lending to them because their bureau footprints are thick, their liabilities footprints are thick, but the fact is that they might have - there might be NTC borrowers, but there are many lenders who shy away from giving loans to NTC borrowers. You know that two wheeler is a business in which 40% of the loan goes to new to credit borrowers. So what happens is that the undiscovered prime concept as well as the Cyclops concept which we talked about, both of them in conjunction actually help us to lend to new to credit with far more confidence right and as we lend and as we build a body of data of NTC credit these scorecards will become more and more stronger, more and more stronger which will allow us to lend with more and more confidence. So in a way we get into a virtuous cycle. So small answer to that question, no I don't think it won't shrink the customer base, it will actually expand the customer base.

Hi this is Shweta from Elara Capital. So my question is on two wheeler finance only. So I always felt that typically two wheeler is sort of an Achilles heel of a particular lending model and larger part of upselling is happening to the two wheeler customer profiles. Of course you had a very elaborate and comprehensive explanations on how we have put up credit guardrails, but nonetheless historically and even from the industry perspective NPAs traditionally have been typically higher for two wheeler business. So how, so if you can provide some color?

Yes so Shweta if you see recall Sanjay's presentation where Sanjay said that by the end of September or actually early October we were almost close to 60% plus prime and when we talk about 60% prime these are customers who are generally bureau score of about 730 plus, majority of them 750 plus or buying a vehicle of 1.5 lakhs plus.

So these are not in the way the sort of the weaker section of the customers. So and the fact is that as our models and as Cyclops matures we expect our prime to go far higher probably you know go 70% plus which means that you are actually funnelling a prime urban customer base into your customer fold. Also what we have done is that which people have which we have not covered here which are very thriving super bike business where we are funding almost close to 600 or 700 super bikes every month with an average price of about Rs. 10 lakhs to Rs. 12 lakhs plus. So overall as a strategy of the organization, our two wheeler business volumes have actually gone up in terms of number of customers as well as ticket sizes plus we are funnelling a large amount of prime and we do believe that this prime customer base which we are funnelling in is significantly different from many of our competitors what they do and again even in prime even in NTC you might have undiscovered prime the concept which Debarag talked about. So our strategy is to get in customers who are prime by their current footprint or their

prime lookalikes and I do believe that as this body of customer grows over a period of time this will form a very strong prime customer base for cross sale of other further products. Yes this won't happen overnight. Right now we are about you know 60% plus prime as we move it to 70% prime and as we traverse through you know we started this journey around September of last year where you will recall in terms of data in September of last year we were roughly about 30% prime. Right now as we speak in October or November of this year we are close to 60% plus prime. So we actually doubled the prime acquisition without dropping our volume. So our sales channels have been oriented that way, we have done a lot of analytics into our dealer channel to figure out which dealer channels actually fit in much more of prime or VPP program actually incentivizes a dealer on prime acquisition.

So even in a storefront where other competitors might be present 100 customers are coming through the door and if the natural distribution of prime in that storefront is 40 of them we will try to garner 30 of those primes, take an outsized share of those primes and obviously as you know primes have much lower loss rates, primes have much more stability and primes lend themselves to cross sell much, much better. So that is sort of the strategy which will play out over the next couple of quarters.

Thank you and congratulations on these digital initiatives.

Shweta just to add I think the next question which comes always whenever this kind of a question is raised is on the returns. The yields naturally are low, but the risk adjusted ROAs will take care of that. So the credit cost, collection cost when it's a new to credit customer and I think the Cyclops is going to be the key differentiator.

When we differentiate the right customer, offer credit to them we are going to have the best of customers with us as well as the ROAs also will not be impacted.

Viral Shah from IIFL Securities. Sudipta I have a question, you guided for 2.8% to 3% ROAs and I think briefly Sachinn touched on it but given the changing book mix, one which is increasing share of urban, second probably reducing share of MFI, thirdly within MFI the weighted average lending rates coming down and fourthly as the leverage going up, how do you anticipate you reaching this ROAs on a sustainable basis by when? Because in last two quarters we have seen that we have reduced the NIMs plus fees guidance by nearly 50 basis points. So can you give us a walkthrough of that?

Yes, I will give the first part of the answer then maybe Sachinn can add to that. One of the things which is there is that obviously you're right you know some portion of our business especially on the MFI business, we are tempering that business currently in line with the market conditions. However, the fact is that it is not that we are not growing other lines of business or the other high-yield lines of business. As you saw we have launched MicroLap which is a secure high-yield business. There are a couple of products which I can't reveal right now which are currently in the works. Some of them might be high-yield. In fact we have launched a small as an experiment what we call the high-yield mortgage also we have launched slightly as an experiment and we are piloting it in a couple of markets in the west of the country. One of the other things is that if you look at our fee profile, our fee profile has been largely dependent on the fees on origination and some parts of CLI insurance on attachment. We are working on a plan to sort of broad base our free profile as well and there are couple of products specific products which can be done which actually adds to your fees profile especially in the line of payments etc. This particular block for us is completely empty. Now obviously we have guided at 2.8% to 3% ROA on a sort of consistent basis. There are a couple of blocks missing blocks which need to get filled in for that to happen. We were at about 2.6% at the end of this quarter. The fact is that if we are getting into a - if you are able to and we have set up a completely a new insurance channel where we actually right now we could do insurance only at the point of origination as CLI. Now we are actually trying to sell through the entire insurance

lifecycle because we have got a corporate agency license. We have put up a fully dedicated insurance sales team as well. So on one hand we are trying to improve our fee profile so addition of new fee based products as well as insurance. We are trying to add a couple of more new product lines which can have higher yield profiles and you will see them as they come along. Personal loans business you will see one launch that will happen tomorrow. I do believe that the personal loans business also has legs to run and it's a relatively medium in product. So overall there is a lot of focus on cost reduction. The lot of digital work that we are doing is an overall focus on cost reduction is what we are carrying and I do believe that from next year probably India will probably start on a downward rate cycle from next year. You have to see when it happens, but we are hopeful that it might happen.

So overall I think we remain moderately confident about NIMs and fees guidance about between 10.5% to about 11.5% percent. As your cost of credit starts weaning off as we have said between 2% to 2.25% is our sort of directional movement towards cost of credit with a bias towards 2%. I think that 2.8% to 3% number delivery in spite of the changing asset mix might be possible on a consistent basis.

Just to add when we said the NIMs plus fee range is being brought down from 10.75 to 11.25 to 10.5 to 11 that was we had also mentioned that this was because of the challenging environment externally. No doubt the regulator has been talking about bringing down the interest rates generally, not just about one business, but across businesses. What will actually happen is the riskiness of doing each and every business is going to come down. So the credit cost requirements at the industry level itself will come down. If the whole industry is not going to take the kind of risk which it takes today it's going to naturally bring down the risks which will lead to lower credit costs. The collection efficiencies will improve across and with the kind of work that we are doing, I am sure that we are going to spearhead that because as the balance sheet starts growing there will be businesses like Home Loan LAP which today the yields are low. But on an ongoing basis as the book starts growing the incremental cost of doing the same business or adding balance sheet is not that high. So the operating efficiencies will then kick in to ensure that the overall opex will go down which includes collection costs and the credit cost of course there's a lot being done on that. So in spite of the NIMs plus fee coming down to a particular level we believe that we'll be able to manage on a sustainable basis 2.8% to 3%.

Can I have one more question?

You can have one more question, but provided others are okay and maybe give others a chance and then come back.

Yes.

So this is Hardik Shah from Goldman Sachs. Firstly congratulations on putting this piece very useful insights on the entire Cyclops, how they are touching each of your businesses. I have two questions one is can you broadly tell what will drive the 20% to 25% growth guidance that you are giving? If you could share what products will drive that? And second question is MFI collection efficiency has seen marginal dip in October, so where is that stress coming from and incrementally where is that heading if you could share in November.

So I'll take the second part of your question first. So I think incrementally as we see in November I think we are sort of equivalent to what we are in October in most of the markets we, actually when I looked at the collection

efficiency numbers as of yesterday, in most of the markets we are actually have an improving bias to those collection efficiencies. We -- you know that we had a flood in the last week of September in Bihar, and that led to some collections disruption in Bihar in the first in October and then it was followed by Chhath in Bihar. So obviously, when our people also go for some leaves etc. So but we are also very hopeful that in Bihar also the collections efficiencies will come back to normal in the latter part of December. In a nutshell November we are seeing similar amounts of collection efficiency in as we saw in October with a bias towards improvement. I am just not getting ahead of myself and trying to guess what December will be, but given the fact that we have had a reasonably good Kharif crop the social flows have started once again, and I do believe that December will probably if not better will probably be around on similar levels. So Sachinn you want to add anything to that?

Yes just I think -- usually we see that whenever flood like situation happens which especially in Bihar, you always find at least once in a year there are certain districts which get into this challenging circumstances, but we have seen that after 60 to 75 days things start coming back to normalcy, so we are very hopeful that come December we should start seeing some improvement that there will be first stability and then some improvement. So you can see certain geographies excluding Bihar already starting to see some kind of improvement. So on an aggregate basis we believe that in Q4 we should be seeing improvement, I mean the collection efficiency whatever reduction has to happen and you have seen in quarter 1 and quarter 2 the overall reduction has been around 40 basis points so, it's not going to be alarmingly high and the reduction which we will see in Q3 will be in about the same range as what we have seen in earlier 2 quarters and post which we are very hopeful that things should start picking up.

loans lot of people are scared about personal loans you have to remember that prime personal loans and super prime personal loans continue to behave properly even after so much of doubts on it. The section of personal loans that does not continue to behave properly is the section of personal loans below Rs. 75,000 of ticket size and in a Rs. 13.5 lakh crores market of personal loans only about Rs. 1.25 lakh crores is below, Rs. 75,000 and that too has a loss rate of about 6% to 7%-odd loss rates while prime personal loans still continue to operate below 2% plus loss rates. Yes, there are questions whether in personal loans there has been leverage etc., and all this. So yes, there has been leverage, but the fact is that the industry has also seen stabilization in risk cost in personal loans and unsecured lending -- in September and October so that business will also continue to grow. I do believe in India mortgage as an asset class will continue to grow at about 20% to 25% growth rates, some secured loans like gold loans, micro lap etc., will continue to grow at 20% to 25% rate. Personal loans, I do believe safe speed to grow is about 20%, micro finance I do believe safe speed to grow is about 20%. So I do believe if you plan your portfolio distribution properly and you plan your growth strategies properly, on a steady state basis growing at 20% to 25% is not a challenge in a safe risk calibrated fashion. Yes, some businesses you might gain some tailwinds, for a couple of quarters you might grow at 30%, but on a steady state basis, we would like to grow at between 20% to 25% with a bias towards 25%. Does that answer your question?

Thank You.

Sir, Ramesh Bhojwani from Mehta & Vakil. First and foremost many heartiest congratulations on 30 years, which we are celebrating today, delayed by 2 days because we started on November 22, 1994 and I have had the good fortune of being virtually in almost every presentation by L&T Finance, beginning with Mr. Devasthali, then Mr.

Dubhashi and now your good self. It's remarkable to see, how L&T Finance has evolved in all the eight verticals of your business, with not only digital strength but with an extra incisive innovative and inventive carefully crafted safety first feature in the lending business which is full marks and it shows that your net NPAs are at 0.86. Sir, 1 Slide was there where it was showing in the last bottom, some write-off of Rs. 256 crores or Rs. 236 crores. Not a question but just would like to get a clarification what is the total write-off and are we recovering even although they are in the write-off segment as banks do?

Yes, I will just take that. So this quarter we had a total write-off of Rs. 641 crores, out of which what you saw was on the rural business loans we had about Rs. 236 crores. The other Rs. 236 crores, it's a coincidence that the number is same is on farm loans and the balance of about Rs. 169 crores was 2-wheeler. Now the write-offs usually happens it's more of a CFO’s decision when to actually take a write-off because once asset is provided 100%, then you can do a write-off on an immediate basis or you can take a write-off at a later date. So all these loans which have been written off, have been written off after 100% provisioning and all the provision 100% provision need not necessarily have come in the current quarters’ credit cost. The credit cost would have already come over a period of time.

And we are certainly pursuing recoveries from this segment, because the classification as a write-off is viewed by RBI as say after 90 days. But they are not bad loans, they may be a business in temporary, intermediary, quarterly, reverse cycle

Absolutely, so especially for rural business loans the moment it is 90 DPD we do a 100% provision. That does not mean and that's why I said that it's a CFO's prerogative, it's a technical write-off.

We have separate collections team following on write-off portfolios and we track recoveries from write-off portfolios separately.

Perfect, it's very well answered. Thank you and all the best.

Hi, this is Nishant from Kotak Securities. So this is a question pertaining to data security and the changes or the regulatory changes that would come in this front. What do you envisage in this and how have you taken care of this in Cyclops?

Yes, so one of the things in Cyclops, obviously we consume Bureau. Bureau obviously is all kosher and regulator approved. We consume account aggregator. Account aggregator obviously comes with the customer consent after the customer gives OTP. On all other fronts we do not consume any raw customer data. We only consume an indicative score which can be an affluence score or which can be a risk score or which can be a stability score.

So we do not take any PII data. So if that answers your question.

Got it. Thank You.

Yes. My name is Vipul Shah. So I wanted to know what is the total capex we have done on this Project Cyclops, this entire digital initiative around which this whole discussion is going on? And also what type of expected returns on this investment?

So, in fact the return, if you had seen slides which Mr. Debarag was showing, he had talked of the term RODI. I think it's a new concept. As far as returns are concerned, we will have to still bide our time to see because it's just being implemented for the first product, first business. It's 2-wheelers. The pilot just happened in June and we started seeing the results which seem to be very good. On the total investments in Cyclops, in the first half we have invested about Rs. 20-odd crores. Another Rs. 25-odd crores will be invested in H2. So about Rs. 45 crores is the investment purely for Cyclops. The total capex that we intend to incur during the current, I mean what's been budgeted is about Rs. 175 crores for the current year. The same number was about Rs. 100 crores to Rs. 125 crores in the previous year. And the rest is all operating expenses.

But what I would like to add to what Sachinn said is that since Cyclops is a product, Cyclops is a standalone product in itself, we are able to capitalise a lot of the cost in line with the accounting guidelines. So the other thing which I would like to say is that Cyclops, one question which arises is that why it is 55% right now? Why it is not fully? You put it in June, beta was released in June. Why you are pushing 55%? Because it was a new product and because you saw that in 2-wheelers we are using 16 scorecards. It's an extremely complex machine. Probably the country has not seen before in credit underwriting. So we wanted to be absolutely certain that what we are doing is correct because the fact is that if something goes wrong then it will show up badly in credit cost. So we actually slowly migrated dealer by dealer, dealer by dealer, region by region and then looked at the results and see whether we need to fine tune anything. So this is a process of fine tuning that this machine is going through.

But as Sanjay said, by end of December, early January 100% of our 2-wheelers will be on Cyclops and then the tractor business will go into Cyclops. And as I said originally in my presentation that this is a productised modular approach where this is an Omni product, Omni credit engine. It can do anything.

Hi, this is Chintan Shah from ISec. Thank you for the detailed presentation and showcasing our digital capabilities especially on the project Cyclops. So most of the efforts seems to be towards the pre-sanction stage which is expanding the customer base and sharpening the underwriting skills to lend to more quality customers. Moreover, now most of our processes are digital in nature. However -- how are we leveraging analytics on an account once it becomes overdue, even if the overdue is by a single day? Can you please share some colour on that?

Yes, Chintan, you probably have seen that couple of Slides on what we call the automated portfolio management engine, which is the next sort of moon-shot that we are taking, right, where it's an automated portfolio management engine. It will, even before the customer goes delinquent, it should ideally tell you that the customer is on the verge of going delinquent. Because the fact is that as I said, credit -- there are two things which I said at the beginning of my presentation. Traditional credit administration is dead. If any organisation in this day and age is following traditional credit administration, they are probably not doing justice to the streams of data and the technological stack that is available in India today, they are not taking full benefit of this. So the way we want to go is that when Nostradamus comes live, which is an automated portfolio management engine stack, which is probably in terms of complexity a programme, which is probably two times more as complex as Cyclops, when it goes live, it will tell us at an individual customer level that this customer is this probability of going delinquent in the next 3 months. And then your entire proactive collections engine will spring up on that customer. This will be done for millions of customers automated in a sentient basis. I use the word sentient, because the machine itself will ingest data sources, the machine itself will run the BREs and algorithms, and then the machine itself will decide whether this customer is behaving, exhibiting risk, and what is the probability of this customer going bad in the next 3 months, which will form a feed to a pre-delinquency measurements. So ideally, I would not like that customer to go to that DPD 1 or DPD 30 stage. I would like to address the customer before he gets there.

I think the date mentioned is Q2FY26.

Yes, so that is what we are saying, Q2FY26, we'll try our best to get there before Q2FY26, we'll try our best.

Thanks.

I think Viral had one more question. Viral, if you want, you can.

Just wanted a perspective on the MFI business that you have been highlighting. As you mentioned today, there's another news that has come out that it will be capped at three lenders. So how do you foresee the second half of the fiscal for especially L&T Finance, given that if we look at that portfolio that is similar to what it is at an industry level of 18%?

It will not be -- it is actually, if you have seen the Slide which Sonia was flashing, it is 8% plus 5%, so it's about 13%. And that 8%, if you looked at the collection efficiencies over there, it's about 97% collection efficiencies. So yes, in fact this recommendation, I was part of the meeting where it was, where these calls were taken. And I think the whole industry, all the members of MFIN are, one thing is very clear to them that if they don't take stringent calls at this point of time, then there are, the challenges are only going to increase. If you are going to continue financing an over-leveraged customer, you are just passing the buck. You are just pushing the problem to a future date. And that's also one of the reasons why they have brought down the overdues, for funding from 89 days to 60 days. So I think at an industry level, there is going to be pain which will increase for sure. Players like us who have faced lesser problems because of the guardrails that we created four years back would continue to have an upper hand. If at all there is any loss which is going to come, I think we have already mentioned the first quarter, the slippages were about Rs. 100 crores. Second quarter, it was Rs. 175 crores slippages. So we have had about Rs. 275 crores which came in the first half. We believe that in the next two quarters, looking at the current scenario, I don't think we will exceed the overall provisions beyond Rs. 1000 crores. So we still have some room and we are just hoping that the way we have been managing this business, we should be in a position to curtail whatever losses are there and if the collection efficiencies improve going forward in Q4 and that's what we believe things should start turning around. Because if the pain level has to come, it's better that people take that hit and go ahead in life because it's not just about the current losses. The customers are there. If you keep providing more and more money to an overdue customer, I don't think the problem is going to get resolved. Yes, you cannot suddenly squeeze out liquidity from these customers, but as Sudipta was mentioning, the external situation should start improving because whatever problems arose was more on account of election, the liquidity drying up because most of the projects, the MNREGA projects and other social projects which are undertaken, they had to be stopped. The liquidity was squeezed out and we see that most of such projects have restarted. In fact, the Ladki Bahin scheme which was announced in Maharashtra, we saw collection efficiencies already improving, although the portfolio is very small in Maharashtra, but these kinds of schemes are being taken up by major states these days and the liquidity is helping people in a way to get back on track. So we are very hopeful that the losses should be limited. If the industry is going through a turmoil, then we cannot completely get out of it absolutely unscathed. What also helps us is that we are sitting on Rs. 975 crore macro-prudential provision and these are the reserves which we have created for these kinds of difficult times. So if there is a need, we will use them.

Yes, but I would like to add to what Sachinn said when I look at data. I see an improvement bias in collection efficiencies in many states. I see an improvement bias. I just don't want to get ahead of myself, but if we see the trend continuing in the month of December, then I think by the end of January, if the same trend continues, we can safely assume that the industry has bottomed out in the December-January months.

Can we take one last question. Kunal first and then we will go to Kaitav so that will be the last question. So Kunal please go ahead.

So firstly, with respect to these entire digital initiatives on the credit side, if we look at it on the existing customer base, how is it mapping? So when you spoke about risk-based segmentation, so how does the existing customer base actually falls within the various risk layers and where eventually we would want to be over a period say next two to three years, if you can just highlight something on that part.

See Kunal, one thing is that it's very difficult for us to go and back-test what we are doing in the current credit engine on the historic base because we do not have those data elements. We don't have those additional data elements. We just have to rely on Bureau and frankly, Bureau gives us a directional sense, but it does not give that fine, sharper separation that we are getting right now. The only thing that I can tell you is that as, for example, Two-Wheeler, the prime base, now throughput is about 70% of our incremental sourcing is Two-Wheeler. So our entire legacy Two-Wheeler portfolio will probably get renewed in the next 12 months Sachinn almost, next 12 months. So I would reckon that after 12 months, my premium prime portfolio in Two-Wheeler will be close to 85%.

Maybe 15% of the legacy portfolio will be left. That will show up in credit costs. So, and as each of this portfolio go in block-by-block into this, all this portfolio will go through the credit renewal or what I would call a much more resilient customer profile. And this is something which is very, very interesting that we saw during COVID. If you look at the Bureau data, before COVID, if you have two sets of customers in the Bureau, the resilient customers and the sensitive customers, sensitive customers defined as those customers who exhibit more risk under stress.

Before COVID, the risk separation between those two customer segments was not more than maybe, 30, 40 max 100 basis points risk separation. COVID comes, the risk separation became 400 basis points between the sensitive borrowers and the resilient borrowers. The objective for doing all this is to make sure that your portfolio has a far greater share of resilient borrowers. You cannot totally eliminate the sensitive borrowers, because the sensitive borrowers at times give you a yield. So, it is very important to calibrate a portfolio mix with a larger majority of resilient borrowers with a small smattering of sensitive borrowers, so that in matters of stress, when stress situations come, or when an industry event come or an industry situation come, the amplitude of your loss does not go beyond a point that you cannot manage. That is the science behind it. For us, it's very difficult to go back and retrofit our existing portfolio with these new tools, because you do not have that granular data. But what I can say with confidence is that 12 to 18 months down the line, once all this goes through, the portfolio that will be left will be an extremely resilient portfolio.

Sure, so not done it even on a sample basis?

We have done it on a sample basis. And basically because we did it on a sample basis we saw that the new portfolio behaves particularly well. Probably I can share offline the results of the sample with you.

Okay, thank you. So, we'll have the last question coming from Kaitav.

Yes, thank you. Kaitav here. Sir, if you can explain in more layman's language, what are the industry first in technology that you're trying to bring out.

Okay, I'm a layman. So, whatever I speak, I'm not a technology professional. So, I understand technology, but I'm not a technology professional. So, I'll try to explain. See, what we are trying to do is a couple of things. First thing that, number one, as I said, we are trying to take a modular approach. What do you mean by modular approach?

Previously, when you go and did development in BFSI, you said, you have to improve UI, UX. So, you do a small Mickey and put that into the stack. Our approach is different. Our approach is that if I'm doing something, I will build it like a box. That means, I will build it ground up. I'll build it solidly like a box so that any business, and I'll build it cutting edge, so that if you want to use this box, whether be it in the UI, UX domain, whether be it in the credit domain, whether be it in the portfolio management domain, whether be it in the service domain, you can

pick up this box, make minimal changes and deploy in another lines of business. So, there's a first sort of philosophy change, right, to move from a disjointed approach to a product modular approach. This is exactly the approach that Silicon Valley works in, product-wise modular approach. This is the approach that they work in. So, that's the first change.

The second change is that we are trying to think GenAI and ML first. Many of the things, for example, you saw the automated attendance that, we are trying to do in ML. Now, I used to get asked this question, what is your attendance in MFI? I said, this is the attendance here. How do you know that the lady is coming and your sales guy is not blindly marking? I came back and thought that I don't know, right, if the sales guy is blindly marking or the lady is genuinely there, then how can we use technology to make sure that this is foolproof? Take a photograph of the FLO with the lady who have come into attendance, go back to the KYC photo, do an image processing algorithm, match them with the ladies of the photo that the FLO has got, where the photo comes with geotagging, photo comes with device, photo comes with a token. So, it is very difficult to forge. And then when I do this thing, it gives automatic attendance. It is tamper-proof. Simple example. But we are trying to think GenAI First. We are trying to think how to use technology first in problem solving. This is what we are trying to use. So, GenAI, ML first. This is what we are trying to do.

The third approach is build versus buy. That is the third approach. Sometimes, when you try to buy, the speed at which the people who make for you is much slower than the speed at which you want to move. Critical applications, you make yourself. Because you have far more control on what you want to do. And that is why we have 300 application engineers. Youngsters, hungry youngsters. And what we have realized is that if you give them tough problems, they really go behind it. So, that is the other main change. Which is basically, you know, where you can build it, you know, build it rather than buy it. And typically, we have noticed that building it is probably cheaper than buying it. And it is much more, you maintain the IP with you.

Which brings to me the fourth point. The fourth thing that we are trying to do is we are trying to build intellectual property. It is not about whether you're trying to do something in technology. But it is about whether what you are trying to do in technology, is it cutting edge? Is it the first in India? Is it the first in the world? You know, our engineers have to ask themselves the same questions. And that is why reorienting the technology teams to problem solutioning so that they build critical intellectual IP. That is why, you know, in our lifestyle index, small solution, but a very effective solution. And we scoured around, we figured out that the solution does not exist. It is non-patented. So, we filed our new patent. So, the fact is that the fourth thing is to build intellectual property.

Overall, why we are doing all this? Because we want to seep in a technology culture within the entire organisation.

We want to make -- because transformation won't happen unless the person in the ground level, the engineer, the graduate engineering trainee who is coming in, the guy in the field who is going to, all are learned to use technology, all are comfortable with technology and all trust technology. So, it has to start from the ground level.

So, overall, we are doing all this to build that cultural transformation within the organisation. Long answer to the question, but I hope I have answered your question.

Yes, thank you.

Thank you, sirs, for the detailed discussion and clarification provided. With these ladies and gentlemen, we conclude this Investor Digital Day. On behalf of L&T Finance, I thank you again for spending your valuable time for this digital and business immersion, and we hope it would have been useful. For any further clarifications, request to contact the Investor Relations team.

*Since the transcript has been derived from a voice recording tool, necessary corrections have been made to remove anomalies as well as manifest but inconsequential factual discrepancies, repetitions in Q&A which would have unintentionally crept in, if any