In this exclusive editorial interaction, Sudipta Roy, Managing Director & CEO, L&T Finance Ltd. (LTF), sheds light on how LTF’s 5-pillar strategy is transforming the organization, with a strong emphasis on its ‘Futuristic Digital Architecture.’ Through revolutionary initiatives like the AI-driven credit engine, ‘Project Cyclops,’ and the innovative D2C app PLANET, LTF is enhancing customer journeys, particularly in rural sectors, while mitigating risks with real-time credit risk assessments. Roy also highlights the significance of strategic collaborations with fintech partners, the power of AI and machine learning in revolutionizing credit evaluation, and how data analytics is shaping future demand forecasting in key sectors like agriculture and tractor financing.
Some edited excerpts:
As the Managing Director and CEO of L&T Finance Ltd., what is your vision and expectation of technology? Can you cite some recent digital initiatives that have made a huge impact on your organization’s growth and profitability?
To drive sustainable and predictable growth, we’ve introduced a comprehensive 5-pillar strategy. One of the key pillars is ‘Implementing Futuristic Digital Architecture’, which aims to revolutionize our digital experiences by streamlining processes and providing exceptional customer journeys. We’ve achieved this by introducing innovative features such as DIY home loan and rural business finance journeys. We have also enhanced our Direct-to-Consumer (D2C) app, PLANET.
For example, we offer a D2C journey to farmers for our Kisan Suvidha loans where no manual assistance or intervention is needed. We’ve launched a DSA channel for personal loans and achieved ISO 22301 certification for business continuity standard practices. Additionally, we have deployed ‘Project Cyclops’, a revolutionary AI-driven, next-gen digital credit engine for real-time credit risk assessment, across selected dealers offering our Two-wheeler Loans. We are aiming to revolutionize the credit assessment space with this innovative digital solution across our other products as well. Our efforts to improve customer experience, application engineering, and robust IT infrastructure are enabling such in-house engineering and accelerating our time-to-market.
At LTF, we are providing better consumer experience with our tech-based offerings. We prioritize secure IT infrastructure to safeguard customer data and ensure business resilience. Our focus is on delivering a seamless, customer-centric experience that fosters loyalty, retention, and growth through digital transformation.
Project Cyclops is one of the most ambitious technology-driven initiatives at L&T Finance, utilizing AI and ML for credit risk assessment. Could you elaborate on how this three-dimensional credit engine is transforming the way creditworthiness is evaluated, and what specific role AI and machine learning play in this process?
‘Project Cyclops’ is a revolutionary ML-based, in-house developed, proprietary credit engine which facilitates an in-depth assessment of the customer’s potential by ingesting bureau, account aggregators and alternate trust signals at scale. The bureau data helps us with the past credit behaviour of the potential customers while the liabilities data through an account aggregator provides insights into the customers liquidity profile. The alternate trust signals indicate affluence, spending patterns or the density of the spending. By transcribing this thoroughly, we are gaining a distilled understanding of the customer than ever before, which will help us underwrite better credit tested customers leading to a superior portfolio quality. We are leveraging these inputs to create machine learning based ensemble scorecards. These scorecards are built after an extensive analysis and overlay of historical credit performance data on all the axes.
The system integrates bureau data, liabilities information, and trust signals from UPI transactions and geolocation data. How is this multi-layered approach providing deeper insights into borrower profiles compared to traditional models, and how does this impact your decision-making process in both rural and urban markets?
Apart from gathering data from three-dimensional sources, we have also enabled a concept called micro-geography. This allows us to identify region-wise risk patterns and apply them to our underwriting process. As a result, ‘Project Cyclops’ is emerging as an omni-credit, omni-product engine that can underwrite all our products like Two-wheeler loans, Personal loans, Home Loans, Farm Equipment Finance among others. Currently, we are using ‘Project Cyclops’ for Two-wheeler Finance as it is a high-velocity business. As compared with traditional models which may underwrite basis 2-3 scorecards, we have built 16 scorecards on ‘Project Cyclops’ for our Two-wheeler Finance business. We intend to extend this to other businesses as well like Personal Loans, Farmer Finance, and SME Finance in the coming quarters. Thus, by combining these data points, we are now in a position to make more informed lending decisions and reduce the risk of defaults across rural and urban markets.
You’ve partnered with external entities. How do these technological collaborations enhance the overall functionality of Project Cyclops, and what is the importance of third-party data in creating a more comprehensive risk assessment model?
We have tied up with multiple fintech partners across various domains to enable us to obtain a comprehensive risk overview of the prospective customer. The ‘Project Cyclops’ engine is a pioneering three-dimensional integrated credit engine in the country that does bureau checks, account aggregators/liabilities, and trust signals in one pass. It is an omni-product that can process any product, with very low latencies quickly and efficiently. This helps us understand the customer and provide them with optimized offers for a seamless experience. With the beta version on our Two-wheeler products, we have understood the architecture and are now actively working on expanding this to our Personal Loans and Farm Loans. We have also filed for trademark registration of ‘Project Cyclops’.
L&T Finance has set ambitious goals for a 25% CAGR growth in its retail book, contingent on rural demand. How is predictive analytics and data from your credit engine helping you forecast this demand, particularly in sectors like agriculture and tractors, where external factors like monsoon patterns play a crucial role?
A good monsoon season is strongly correlated with increased credit demand and overall per capita growth. When a monsoon season is favourable, it leads to improved crop yields, resulting in increased agricultural income for farmers. This, in turn, increases their demand for tractor financing and other credit products. Additionally, as farmers’ incomes rise, they are more likely to spend on consumer goods and services, which further fuels economic growth. At LTF, we leverage our partnerships with multiple fintechs to analyse the correlations between average rainfall, crop yields, reservoir levels, and soil tides that helps us mitigate risks
Our portfolio-mix between rural and urban is well diversified with 48% and 52% respectively. Going forward, our focus will continue to be on garnering prime customers across both the segments and move towards predictable portfolio performance. We have already achieved our Lakshya 2026 goals two years in advance at the retail level and we aim to achieve and maintain the momentum of other parameters too.
In the next six months to one year, what are some of the high-impact technology initiatives that your organization plans to undertake? What will be the impact of these initiatives?
LTF is exhibiting its digital prowess with an introduction of new initiatives like the futuristic Lifestyle Index Calculator, that revolutionises the way customer data is collected during field visits. We have even filed a patent on Lifestyle Index. The AI teams are also writing paper on Lifestyle Index on international journals.
We are also establishing simplified and frictionless loan journeys of our PLANET App, and are extending ‘Project Cyclops’ to other businesses, which is presently launched for Two-wheeler financing. The organisation is spending heavily on technology infrastructure, technical resources, and training of existing manpower to undertake high impact initiatives. In fact, our Bengaluru centre has been set for this purpose only. We are also planning an initiative where we are showcasing our AI expertise and connecting with industry experts. The event is called R.AI.SE and it will be held later this year. There, we are calling speakers from Silicon Valley and top companies across the world. Besides, we are also working on couple of Proof of Concepts in Generative AI using Google multi-modal LLM (Large Language Model) Gemini and some of these will go live in Q3FY25.
In the near-term, we will continue to focus on enhancement and optimization of our talent pool, especially in the fields of AI and ML, Credit & Risk, as well as Tech & Engineering.