PayNearby’s journey of democratizing credit disbursement in Bharat

India, today, is undoubtedly one of the world’s largest fintech centres. As per the latest report of BLinC Investment Management, the overall market size of the fintech industry has grown to a figure of USD 31 billion [1]. It is estimated that India has more than 6,000 fintechs, of which 16% are involved in lending [2]. With the exponential rise in UPI-driven digital payments, use of the internet and e-commerce, growing volume of assisted banking at retail stores, data and Artificial Intelligence (AI) have the potential to revolutionize the availability of credit to the missing middle and lower-middle segment in the country.

The evolving fintech domain has given hope to the Indian SMBs looking to transform their local businesses and attract online customers. Retailers often need timely and small-ticket working capital loans for short tenures to keep their businesses up and running. More often than not, due to lack of collateral or being new-to-credit, these retailers predominantly rely on alternative options such as local moneylenders or loan sharks instead of mainstream options such as banks or financial institutions. But, loans from these informal sources come at exorbitantly high-interest rates, pushing them further into an inevitable vicious cycle. This is where data and AI come in.

Data is actually the first step to creating a digital lending ecosystem. Data is generated at multiple places including our bank statement, shopping habits, GST return or statements, lifestyle habits including cab usage, watching movies, apps we use for our day-to-day needs etc. For example, when a small retailer sells or buys something digitally, the footprint of that transaction gets generated. Invoice details of digital purchases are uploaded automatically to the GST portal to file taxes. Therefore, as more consumers and small businesses transact digitally, they create digital footprints. However, as this data is available at different places and in different formats, how can we aggregate this data and get some meaningful insights out of it?

Knock! Knock! - AI is here to help us all!

The greatest advantage of AI is its ability to handle unstructured data, such as images, videos, geographic locations, excel or word data, trends, and transactions. AI can help analyze the problems faced by the lending industry and help in:

• Faster and improved onboarding process
• Better credit underwriting and decisioning
• Faster decision-making
• Improved risk management techniques
• Reduced underwriting cost
• Fraud detection
• Enhanced security and compliance
• Credit monitoring
• Debt recovery

The bank statement analyzer technology, for instance, is built on the premise of AI/ML, and it evaluates the operational data (i.e. debit and credit transactions, transfers and end of the day balances) from a consumer’s bank statement and converts it into insightful data which can be customized for each lender. This helps evaluate applications at a great speed, keeps the costs low, and a higher number of applications can be processed every day.

We, at PayNearby, started our lending vertical amidst the lockdown in April 2021, to provide access to credit to our microentrepreneurs. The biggest challenge faced by our retail partners was the lack of formal records of income and expenditure, business track record, nil or improper credit history, and a very high cost of last-mile credit availability.

To address the above issues, we developed a PayNearby ‘Bharosa’ Score which evaluates each retailer on a scale of 0-10 and makes them eligible for the lending services offered on our platform.

This score is AI-based and evaluates more than 100 parameters that provide insights into the behaviour of each retail partner. Each eligible retailer, based on the above score, gets an opportunity to apply for loans with our Lending Partners.

Our motto is customer delight, and hence we have automated the entire loan origination process, including review of documents, credit bureau checks, etc. This has led to a lot of cost optimization and the removal of any bias in advance while processing a loan application. Moreover, integration with partners for real-time KYC has resulted in faster application processing and improved TAT, and the model is highly scalable. Authenticating customer profile against government data sources in real-time makes the application journey smooth, fast and error-free. Through partner integrations, the data is transferred immediately to our lending partners who, in turn, can process the applications faster and hence, disburse a higher number of applications in a day. We started with this service in some select states in India but, within six months of operations, this service is now accessible all over the country.

We have employed AI and ML to not only source loan applications digitally on our tech platform, but also to assist us in comprehending the creditworthiness of applicants. This, in turn, has allowed us to offer the right loan amounts to the applicants. To date, we have processed more than 3000 applications on our platform and have established a great use case for a data-led digital lending model at PayNearby. We are committed to fortifying our retailers with financial backing to ensure their businesses function smoothly by bringing them into the mainstream credit fold. Our innovative and robust technology-backed platform reaches out to the deep roots of the country and creates simple and efficient micro-lending experiences for our retailers, enabling them to secure loans at the click of a button.

Sunita Manwani, Business Head - Lending, PayNearby