Credit scoring companies have to invest in tech and price their products right: CRIF India
In this segment of BFSI Tech Tales, Pinkesh Ambvat, CIO & IT Director, CRIF India, shares the technology strategy at the credit bureau, and how they view failure and innovations in the technology roadmap.
Credit scoring companies consume and analyse enormous amounts of data and to ensure that they remain competitive they've to deploy and invest in technology effectively to price their products in a competitive market. Pinkesh Ambvat, CIO & IT Director, CRIF India, in this segment of BFSI Tech Tales shares the strategy on how they've built their platforms since inception in India and what are their plans in adopting emerging technologies, cloud adoption and shares some points on failure as well. Edited Excerpts:
At CRIF High Mark, Pinkesh shares, services provided to Indian customers (financial institutions) are developed in India as compared to their competitors. The core product is the credit bureau platform which is their bread and butter where they provide credit reports and credit scores to their customers (banks, financial institutions and retail customers as well via B2C channel).
Explaining the technology stack, Pinkesh shares, “The main pillars of our technology stack are Java and Oracle, our journey started in 2010 when we were only providing services to Microfinance companies and then entered into consumer segment and commercial segment thereon. In the last three years, we’ve started working on different sets of emerging technologies from Big Data to Automation.” The volume of data in the credit scoring space has grown exponentially as industry gets connected more closely than before. He adds, previously in 2010 we used to receive 50,000 enquires in a day and right now we’re receiving 1.5 million enquires in a day, and all of these are real time enquiries. Apart from this, we receive monthly transaction data from financial institutions. He says, “To process this kind of humongous volume of data we need to have robust systems. We cannot simply increase the cost by investing in costly software and we adopt open source technology like Hadoop etc”. Pinkesh adds that for most of the batch processes they leverage open source technology and for real time processes Java and Oracle platforms are leveraged.
Cloud Adoption & Emerging Tech
Cloud adoption gives flexibility to an enterprise to scale up and is also cost effective. He says, “We’ve been using cloud for the past 2-3 years for different products. We are using cloud for DR strategy as well. DR is a dead investment; it's used when only a disaster occurs now with cloud adoption. I have to be ready with at least 50% of my infrastructure in DR. With cloud adoption I can start very small and I can scale up whenever disaster happens as I don’t require having all investments upfront.” They’ve been using services from different cloud providers for different products which are at different stages. He adds, “There are a lot of benefits with cloud adoption, as we provide many of our products in a SaaS model (Software-as-a-Service), this enables us for faster onboarding of clients (lenders) and it becomes very easy, cost-effective and not much resources are spent.”
“It’s a journey and we would continue to explore areas where we can adopt cloud as the future lies there.” On leveraging the emerging technologies like Big Data Analytics, Artificial Intelligence, Machine Learning, Ambavat says: “We are living into the production stage from the last one to two years. We have completed phase 1 and in phase 2 we are pushing adoption of these technologies into not only our processes but products as well.” According to Pinkesh, getting the right talent for building emerging tech capabilities is not a big challenge. The strategy they’ve adopted is by continuous training and the hiring policy has been kept as for every two new employees they hire, five employees are trained and reskilled.“ We have done a lot of investment in retooling and reskilling the existing workforce who’ve been long associated with us and it’s a motivational factor.”
Integration with New Age Lenders
Integration has always been a challenge for financial institutions as complex systems need to talk to each other seamlessly. On the point of over integration challenges with new age lenders, Ambavat says that they are learning from the new age lenders as well as they’re more demanding in terms of technology. “For e.g. Previously we used to deliver XML as standard output now they are asking for JSON. They’re asking for SLAs in seconds, uninterrupted 24*7 services. They’re giving us a chance to be flexible and adopt new technology as well.”
Failure & Innovations
In this competitive (credit scoring) space credit scoring companies have to rightly price their products and accordingly invest and maintain their tech platforms. Pinkesh adds, “To survive in this space, technology investments have to be kept at check and maximize the benefits out of it in a smart manner.” On an innovation which didn’t scale up as envisaged, he recalls, “We started our journey with big data and we were not able to find the right kind of people in the market to go ahead with this. We started learning by ourselves and a few POCs (Proof-of-concept) which were successful decided to go on the production stage and migrate one of our products on big data. One year later we didn’t see any results and weren’t able to do it and we couldn’t find the right partner either as there are really a handful of people in the market who command knowledge and expertise in big data on production and how to manage it.”
Followed by few changes in strategies, we were able to migrate 60% of products into big data analytics and have managed to reduce cost tremendously and process more volumes of data, he concludes.
Source: Publication: BFSI , 6th Oct,2020