How Edelweiss Retail Finance used AI and ML to its advantage in the times of COVID-19

The SME sector has emerged as one of the critical building blocks of the economy however over the years, the funding to this segment has remained largely stagnant by the mainstream banking sector. To fill the gap, many NBFCs and intermediary financial institutions have emerged to finance SMEs. The role of digital and emerging technologies is playing a key role in financing these micro, small and medium enterprises in a cost effective way. Companies like Edelweiss Retail Finance are using AI and the power of low code systems to empower SMEs.

Edelweiss Retail Finance disburses over fifteen percent of the loans based on AI capabilities. The company has also recently begun working on certain propensity models based on data and marketing campaigns.

AI and ML was leveraged in doing risk analytics of the existing customers, in the background of COVID-19. “We were able to cohort our customers based on their payment patterns in response to the moratorium announcement. The customers were divided into red, amber and green,” says Mehernosh Tata, CEO, Edelweiss Retail Finance Limited. The green fall into the category who are the most likely to pay their next installments. The red would be the least likely and the amber are in the middle. The historical past performance data was also looked at. The data models were prepared, which predicted the customers who are likely to drop on paying back post the moratorium completion. “We proactively reached out to them to explain the importance of not bouncing and maintaining their credit history, and as a result about ~25 percent of the customers marked in the red zone did pay their respective EMIs for the months of Sept-Oct,” informs Tata. The AI and ML algorithms are finetuned on a regular basis for the accuracy to be consistent.

Even prior to COVID-19, based on Data analytics and AI patterns (based on past behaviour), Edelweiss came up with a pre-approved product for an existing cohort of customers. The checks and balances for these customers were already done based on these well paying customers and when they were reached out for a loan top-up, it was well received by them. The processing time for disbursing these loans was reduced from seven to two days. This was because the processes to be followed, hitherto when these customers would approach us for a loan top-up were already completed beforehand.

The Goods and Services Tax (GST) data also throws many markers that can be used to devise a financial product. The current products in the same range are based on the combo of surrogate banking and GST. “We are planning to launch a financial product, purely based on GST data. The necessary data gathering exercise is in process. AI and ML will play an important role in composing this product,” says Tata.

Edelweiss Retail is a digital immigrant company and runs their loan origination system (LOS) on a low code platform. The company is also looking for data robots, which will tabulate data, devise AI and ML models and also run it in Live. The API gateways have been created to fetch data from the necessary platforms which are openly available for access – PAN, Aadhaar, ITR, etc.

If you have an interesting article / experience / case study to share, please get in touch with us at [email protected]

Advertisement