AI adoption across banks and Non-Banking Financial Companies (NBFCs) has shifted from experimental pilots to a core infrastructure layer, driven by the need to slash secured lending timelines.
While a typical secured loan file takes 18 to 21 days to move from sourcing to sanction, the bulk of this delay stems from administrative coordination rather than underwriting.
Rajat Deshpande, CEO and Co-Founder of credit infrastructure platform FinBox, said the sector is experiencing its sharpest technology pull at the origination stage.
Lenders are aggressively deploying AI to eliminate the “file assembly tax”—the hour credit managers waste organizing raw documents—and to fix the “bounce-back loop,” where files shuffle between relationship managers and operations teams multiple times due to missing data, he said.
Removing Forms
FinBox has already deployed its Atlas platform across five financial institutions, with advanced modules for credit appraisal, fraud detection and institutional configuration currently in active pipeline development. It works within the clients eco-system to ensure data confidentiality.
As digital lending scales, priorities are moving from fixing user interfaces to removing forms entirely. This shift enables conversational onboarding through chat, voice, and video, making credit accessible to tier-II markets and small businesses. Concurrently, fraud monitoring is moving from a post-sanction check to real-time, “in-journey” detection that catches document tampering during upload said Deshpande.
On compliance, AI agents ensure automated audit trails by running deterministic policy scripts. Driven by these shifts, AI infrastructure services are projected to be a primary revenue catalyst through FY27.
Native Dialect Converting
Traditional digital applications rely heavily on forms written in English. Over half of India’s small businesses operate in tier-2 cities and rural areas, where these assumptions fail.
The platform adopts conversational onboarding to remove the language barrier entirely. Users can interact in their native dialect using voice, video, or chat. This approach opens up formal credit systems to an entirely new demographic of borrowers, said Deshpande.
The true value of conversational AI lies in its ability to translate informal inputs into structured data. While a borrower shares business details via a casual voice message or chat, the AI works in the background to extract relevant financial metrics and personal data. It organises raw inputs into a clean, decision-ready file for underwriters.
Published on May 25, 2026