‘In the next 12 to 18 months, we want to use data intelligence to further build our core capabilities’

How does data science add value to PayU credit’s business?
The explosion in data has given an opportunity to revolutionise digital lending, making data science an essential part of PayU’s credit business. As a core of our solution, data science driven by tools such as big data analytics, AI and ML algorithms drive multifold value. It helps us manage and analyse risks and at the same time helps us improve customer experience by lending to those who are under-banked. With the help of data science and robust data platform technologies, we responsibly build insights on consumers which will eventually also help our merchants and the ecosystem to grow their businesses.

As part of our commitment to data science, we introduced Project Next last year, which is focusing on building our data science and data engineering teams. This will help us in building a machine learning platform with capabilities to standardise data science, scale-up use of alternate data sources and launch new models rapidly, which can take PayU to the next growth trajectory.

How is PayU Credit leveraging emerging technologies such as analytics, AI, and ML for data analysis to mitigate credit risk?
Over the last few years, risk management has become our core competency at PayU Credit. When I joined PayU in December last year, PayU operated through four businesses, mainly – payment gateway, PaySense, LazyPay, and Wibmo. When the business developed into two dimensions of payments and credit, it scaled up with the data-power of each of the businesses that were a leader in its space.

To mitigate credit risk, we implement complex, data-centric AI-based Machine Learning (ML) models to determine the ‘worthiness’ of a customer – both their ability to pay as well as their willingness to pay. We have built technology based on cutting edge decision models that work on alternate data sets like customer payment information, spending patterns, banking and networking data. We have systematically tested our alternate models and have achieved best in class results.

We often take a ‘low and grow’ approach, starting our relationship with the customer through a deferred payment experience of just a few hundred rupees. We keep growing the loan amount as we see timely repayments from these customers. As the loan amounts get larger we may look to verify incomes, obligations etc using more traditional banking means to offset risk. This approach allows us to digitally serve a large credit-worthy population in India without the need for excessive paperwork and offline processes. Data science helps us scale a customer’s journey from deferred payments to big-ticket personal loans.

Which technology is PayU looking to leverage in the near future with data science to mitigate credit risk?
In the next 12 to 18 months, we want to use data intelligence to build a moat around business and further build our core capabilities. We aim to bring synergies across data and talent across payments and credit business.

As an example, if our credit customer is missing their loan repayments, but is spending on vacations, which is something we know from our payments data, clearly the customer has the ability to pay, but not the willingness. Clearly, a customer which our credit collections team will pursue. Similarly, being able to determine the affluence of a customer from our payments data, we can create a large segment of customers who are pre-approved for a certain credit limit.

In addition to cross leveraging our data and team, we are also exploring newer AI techniques such as knowledge graphs and embeddings to better predict credit risk. Lastly, we are also looking to use what is called MLOps for better end-to-end management of ML models from development, deployment and monitoring.

The next big game-changer technology in the digital lending space?
According to Gartner, by 2021, Artificial Intelligence (AI) augmentation will create $2.9 trillion of business value and 6.2 billion hours of worker productivity globally. Additionally, the global business value of AI in finance will be $300bn by 2030 with 77% of the world’s population using an AI-powered device.
Therefore, due to the extensive presence of AI in everyone’s life, our key priority will be to scale up on AI both for our credit business and payment. In the next few years, AI will be as disruptive as the internet has been in the last 20 years. Therefore, for an emerging digital-led country like India, where credits are thin, it is more important to understand the needs of the customers and merchants and react accordingly. As I mentioned earlier, we have begun experimenting with cutting-edge AI technologies and will continue to do as AI algorithms evolve.

How has PayU leveraged data science to enhance customer experience on their platform?
Through AI and ML, we want to provide the best in class services to our customers and merchants. We want to use AI and ML to remove the risks to enable seamless access to credit, connect businesses to consumers, and onboard new customers digitally. We are using AI and ML to prevent fraud detection and securing data. This helps in peace of mind for customers and merchants, and thereby adds value to their user experience.

Artificial Intelligencecreditdata analyticsData Sciencemachine-learningPayUrisk managementSachin Garg
Comments (0)
Add Comment