How L&T Finance shifted gears from traditional lending to an AI-driven one

Ever thought your daily expenses and geo-spatial affluence around home could decide your loan status? That is what L&T Finance is doing to enable differentiated and personalised offerings, drastically expediting loan disbursement. 

L&T Finance has launched Cyclops, an in-house developed, artificial intelligence (AI)-driven digital credit engine, determining the repayment capability and credit quality of potential customers.

The new model allows for the use of non-traditional trust signals like expenses as a proxy for income via payment aggregators, geolocation and geospatial affinity to enable differentiated and personalised offerings. L&T Finance is the winner this year in the Best in Innovation (Large NBFCs) category in the BT-KPMG Survey of India’s Best Banks and NBFCs.

According to the company, Cyclops analyses customer expenses as a proxy for income, primarily sourced through payment aggregators, and the engine synthesises geo-spatial affluence indicators from various data sources around the applicant's home address.

For the farm loans, the engine uses specialised agri models constructed down to the village level. Models integrate multiple data streams such as current multispectral satellite images, meteorological data, and traditional land record information, creating a unique risk profile of the rural economy.

The impact of this AI -driven innovation is already visible, with sanction time reduced to less than two minutes for the two-wheeler loans. It also shows lower delinquency trends at 83% as compared to the industry, indexed at 100% for two-wheelers.

The engine has crossed 7,50,000 cases dispersed in the two-wheeler vertical and more than 21,000 cases in the farm equipment finance segment.

“At L&T Finance, we have pivoted from traditional lending to an AI-driven, risk-first architecture powered by Project Cyclops, our proprietary next-gen credit engine. Project Cyclops represents a fundamental shift in our business model, functioning as an autonomous real-time decision layer that eliminates subjective bias through end-to-end automation,” says Debarag Banerjee, Chief AI & Data Officer, L&T Finance Ltd.

He further explains that what differentiates Project Cyclops globally is its sophisticated data ingestion and decisioning algorithms.

“By converting unstructured field data into structured, machine-consumable intelligence, leveraging machine learning based ensemble scorecards, and determining the right offers for the right borrowers based on such precise segmentation, allow us to finalise disbursement decisions in seconds with surgical precision,” says Banerjee.

As of Q3FY26, Project Cyclops has proven its immense scalability with 100% uptime, successfully handling high-volume concurrent logins across business segments like two-wheelers, farm, SME, and personal loans.

It is also a winner in the Best Talent and Workforce category for using tech as an enabler to drive employee experience, innovation in talent management and talent acquisition practices and initiatives around diversity, equity, and inclusion (DEI) and governance.

“At L&T Finance, our journey toward becoming a risk-first, tech-first company, is fuelled by a fundamental shift in how we cultivate human capital. Our talent strategy is deeply intentional and is precisely aligned with our long-term Lakshya goals, ensuring every intervention drives organisational success,” said Nilesh Dange, Chief Human Resources Officer, L&T Finance Ltd.

The company overhauled the acquisition engine to prioritise AI-fluent campus talent and implement rigorous, merit-based metrics. Central to this transformation is the company’s C.U.R.V.E. of Happiness philosophy, ensuring every team member feels Connected, Understood, Respected, Valued, and Enabled.

“Through women-friendly policies and a steadfast commitment to work-life harmony, we have cultivated a workplace where trust is the bedrock of performance,” he added.

@richajourno