Aye Finance Pilots GenAI for Image-Based Underwriting of Trading Businesses

Aye Finance, an NBFC backed by Google Capital, is pioneering the use of Multimodal Large Language Models (MLLMs) to solve a chronic problem in emerging market finance: the lack of formal financial documentation for micro-enterprises. By using store images to estimate monthly sales, the company is effectively digitizing the visual inspection process that has traditionally been manual, subjective, and expensive. This pilot marks a significant shift from traditional credit scoring to AI-driven predictive underwriting, specifically targeting the grassroots trading sector in India's Tier 2 and beyond cities.

The strategic context of this development centers on the cost-to-serve, which remains the primary barrier for lenders targeting the micro-MSME segment. For a lender like Aye Finance, sending a human agent to physically verify a small grocery or garment store's turnover is often prohibitively expensive relative to the loan size. By automating this via GenAI and Machine Learning, Aye Finance is optimizing its unit economics. This move builds on a foundation laid in 2019 when the company established its Data Science and AI Unit, signaling a long-term commitment to tech-first lending rather than a reactionary adoption of the latest AI trends.

“

Aye Finance, an NBFC backed by Google Capital, is pioneering the use of Multimodal Large Language Models (MLLMs) to solve a chronic problem in emerging market finance: the lack of formal financial documentation for micro-enterprises.

The technology itself represents a sophisticated application of Multimodal Large Language Models. Unlike standard LLMs that process text, MLLMs can interpret visual data—in this case, store premises—and correlate it with proprietary internal datasets to generate financial insights. This reduces reliance on subjective individual judgment, which often leads to inconsistent lending decisions. By standardizing the income estimation process, Aye Finance not only speeds up the application lifecycle but also ensures a fairer distribution of credit based on objective visual evidence of business activity.