Koushik Kadidal, Chief Data Officer and Head of Insights Business, PayU India, dives deep into how PayU utilises cutting-edge technologies like AI, ML, and data analytics to transform its payment and credit businesses. This interaction explores how these technologies are used to improve customer experience, make informed decisions, and navigate the challenges of large-scale adoption.
As a leader in digital financial services, how is PayU leveraging new-age technologies such as AI, ML, etc.? Can you elaborate on some specific areas where PayU is utilising these technologies to improve its services and offerings?
Technologies like AI, ML, etc., were always cornerstones of PayU’s innovations and evolution, and they continue to be integral today. We are deeply committed to leveraging AI, ML and even data, responsibly and ethically. In fact, they are crucial to our credit and payments business, where they help by delivering business insights, controlling risks, and achieving customer satisfaction. FY25 is going to be a game-changer as we accelerate the adoption of AI, ML, and data across our businesses.
For our payment business, we leverage AI, ML, and data to perform various critical processes for ourselves and our customers. For instance, we enable thousands of our merchants to boost their payment checkout conversions through our AI-powered recommendation engine. This engine analyses up to 10 parameters to prioritise the preferred payment options for the customer, boosting the success rate by up to 5% and reducing drop-offs at checkout. Within PayU to get relevant insights, our team has developed many AI-and data-powered models to verify website completeness of merchants, identify their LOB, and flag high risk websites that may be violating PayU’s policy. Such models help us mitigate business risk, while also decreasing the costs associated with manual reviews.
Similarly, in our credit business, the entire customer lifecycle and experience is based on AI, ML, and data use. For example, these technologies simplify the process of determining a consumer’s credit risk, simplifying customer onboarding when applying for a loan, enhancing customer retention and reducing churn, and determining the need for different lending products and increasing cross-sales of products.
Also, for Wibmo, we leverage a combination of rule-based and ML-based systems to detect fraudulent transactions. Rule-based systems help us detect specific patterns. Say, when a merchant exceeds a certain money limit on daily transactions, it is paused for further transactions until an expert looks at these transactions. Whereas ML-based systems are intelligent systems that auto-detect any anomaly in transaction patterns. For instance, if there is a high frequency of transactions within a short amount of time for a given merchant.
Tell us more about how PayU is using data analytics to make informed strategic decision-making across the organisation? Can you share an example of any successful data-driven initiative?
As I mentioned, we have a symbiotic relationship with data. At PayU, we recognise that data holds a transformative power that effectively helps translate our commitment to providing a secure and seamless environment for our customers, for them to thrive and stay ahead of the curve. Apart from the previously shared examples, another use-case of how we leverage data insights to drive business goals and targets for PayU is by focusing on nurturing the right merchant segments, especially the SMB merchants. Our internal teams have developed a model that predicts the life-time value (LTV) of SMB merchants. This model factors in the growth potential of the merchants along with their likelihood to churn, exhibit risky behaviour, and contribute to high operational cost. This foresight, facilitated by our robust data analysis, helps us identify and serve the relevant merchants, which in turn also translates into savings for us and them.
What are some of the biggest challenges PayU faces when it comes to utilising AI and data analytics effectively? How are you working to overcome these challenges?
It is important to understand that at a pilot level, the application and deployment of AI solutions is relatively easy along with the benefit it generates. Scaling these solutions enterprise-wide adoption is significantly complex and requires more than just a precision solution. These requirements entail: 1) Scaled infrastructure to enable the cost-efficient real-time processing of big data, 2) Adequate governance measures to ensure quality of solutions, fairness, and compliance, 3) Persistent and strong leadership to champion the adoption and integration of AI solutions. These three challenges often create bottlenecks in achieving organisation-wide data and AI adoption.
At PayU, we are addressing these bottlenecks by fostering leadership commitment, investing in robust scalable infrastructure and most importantly, implementing the highest standard of governance frameworks while working with AI, ML and data.
How is PayU ensuring responsible data collection and usage, especially in a region like India with evolving data privacy regulations?
At PayU, we recognise that data is not just any information set, it’s a powerful tool that transforms the way we serve our customers. Therefore, as one of the leading financial services providers, we ensure that compliance is at the forefront of all our businesses. We are extremely committed to using data and new technologies such as AI, ML, in a responsible and ethical manner. Enabling our merchants with a secure and seamless payments platform is our promise, and hence, all our products and operations are supported by well-defined governance models and responsible AI frameworks. In fact, to ensure privacy and security are deeply embedded across our businesses, we’ve created PayU’s privacy and security-by-design policy and toolkit, which is backed by appropriate global training and awareness processes.
How is PayU’s focus on AI and data analytics impacting the customer experience?
We recognise that every interaction with a customer is an opportunity to enhance satisfaction, and therefore, we leverage data to drive improvements that are grounded in facts and thorough analysis. Along with predictive analytics and data, we incorporate AI, ML, and other reinforcement-based learning models within our business. In fact, by harnessing these relevant insights, PayU enables its merchants with business-critical information and proof points such as purchase information, customer buying insights, sales reports, inventory forecasts and other insights. This, in turn, helps merchants to scale their businesses quickly. Additionally, frictionless merchant onboarding, servicing automation, etc., are common use cases where we use AI, ML to improve customer experience.
Looking ahead, what role do you see AI and data analytics playing in the future of PayU? Are there any specific areas where you see PayU making significant advancements?
We are bullish about the impact of AI, ML, and data on our business growth and, in fact, on the overall trajectory of the financial services industry in India. At PayU, we’ll continue to focus on building our core capabilities, including utilising the latest LLM/GenAI technologies, to take full advantage of these advanced technologies. We are also committed to finding ways to bring synergies in data and talent that will propel us to develop, refresh and deploy new models, with proper guardrails, to deliver best-in-class payments and credit experience to merchants, banks, and customers.
According to you, what is the scope of GenAI and data in fintech?
The possibilities are endless. There’s no doubt that in the coming times, we’ll see deeper integration of GenAI, data, and more such technologies into the financial services sector. Along with advancements in several key areas such as fraud detection and prevention, credit scoring and risk assessment, etc., we can expect AI to introduce new innovations in many other facets of the fintech ecosystem.
One such area I foresee is ensuring regulatory compliance. As the regulatory environment continues to evolve, fintechs will increase their reliance on AI and data to automate regulatory reporting into various day-to-day processes, making the reporting more streamlined, efficient and error-free. Additionally, we’ll witness more personalised experiences and solutions being offered to the customers.
Another use case that would come in is to use large language modules (LLMs) to enhance data cataloguing and querying. This will help further democratise the use of data for organisations.