Cholamandalam, a 50,000 crore investment and finance company offers vehicle finance, home loans, home equity loans, SME loans, investment advisory services, stock broking and a variety of other financial services to customers. As microfinance, home equity and the whole line of credit remains crucial to the company, it extensively relies on technologies such as analytics and machine learning to drive efficiency at work.
As the Head of Analytics and BI at Cholamandalam, Vishal Ahuja deals with the analytical approach, BI tools and provides analytical and data management support for all lines of business. Analytics India Magazine caught up with Ahuja to understand the innovation he is bringing using analytics and machine learning.
Ahuja confesses on Chola being an analytics hungry organisation with everyone from top to down are driven by analytics. “Chola is analytical. There is always a push on how to permeate analytics in every facet of the business. Lines of business are using analytics to maximize the marginal return of investment across products, Functions e.g. Treasury is using to forecast the inflow and outflow of money from Chola system” he said.
Analytics India Magazine: What are the key initiatives you drive as the head of analytics at Cholamandalam?
Vishal Ahuja: We at Cholamandalam receive more than 60,000 applications every month for the various services that we offer in terms of loans and equity. All these applications are subject to approval and decline and it is often not clear on what grounds should it be approved. To ensure that it happens smoothly is where my focus area is. Me and my team work on data collection, strategies for business intelligence, data-driven predictive modelling and finding what kind of incentive schemes are working for us, what is doing good, what is not, which employee and brand are doing a good business, and so on. All the elasticity such as price elasticity with respect to a promotion scheme, predicting if the sales would go up and for other areas, the insights are supported by my team.
AIM: Would you like to highlight some of the use cases where analytics has helped you transform the business?
VA: There are two very good use cases. We are making risk-based pricing engine (RPE) for the company. Earlier, pricing used to be driven by market forces. Now, we use a risk-based engine driven by machine learning algorithms to understand how to separate the risk of a market from individual applications. Earlier every customer would get the same price but now the pricing differs based on the risk profile of the individual customer. It is a beautiful use case where we are using advanced analytics and machine learning to get the right price for the customer.
The second use case is the early warning system (EWS). For home loan and home equity business, the ticket size is usually quite large. Late stage delinquency resolution using various acts like Sarfaesi, Legal etc are too time-consuming and even though the collateral value will mitigate losses however impact to company’s ECL will be significant when account moves to stage 3. To have this controlled we built EWS for our home equity and home loan businesses. This system uses alerts from the bureau, market data and internal performance data[v1] which are in turn built on the rules. These indicators are used to check whether the accounts are going into delinquency in the future. The system generates alerts for customers at high risk and these alerts are passed to field collections team to keep a hawk eye.
AIM: How is the company using analytics to approve loans or to analyse the customers?
VA: As mentioned earlier, we use an underwriting engine. What happens is, our executive meets the customer and enter all the application fields to collect data. This data along with the other data which we already have will hit the ML Azure cloud, which has our underwritten logic. Then we get the API call to check whether the account is approved or declined.
We have roughly more than 1000 attributes on each application to get the profile of the customer to check whether the customer is worthy of a loan or not. It is a huge comprehensive exercise that we have done here to ensure the best results.
AIM: What are the analytics tools and machine learning tools that are being used at Chola?
A: We are using Python, R for our analytics engine, and QlikSense for our visualisation. We use all open source tools and have not made investments on the analytics tools.
Q: How big is your analytics team?
A: We are now a 16-member team and we are hiring for more positions in analytics and machine learning.
Q: How has been your analytics journey so far?
A: I have always worked in an MNC environment. After graduation, I moved to GE Capital in Bengaluru and have since then worked in HSBC, Genpact and American Express in the US. From there, coming to an Indian firm was an eye-opener. There is so much to do here. The whole scene India is graduating so fast to maturity. All this information and the richness of the bureau availability needs to be stitched together for which we need more people in the analytics domain to convert the businesses when it was done in an old school way to newer approach. We absolutely need the change or we won’t be in this game 10 years from now.
Each and every person in our team has been hired by me and are extremely passionate. They come from companies like MuSigma, Fractal, American Express, premier institutes like IITs, ISI, DSE. At Chola, we encourage free willing environment and we have a lot of budgets to drive analytical capabilities and that is where we are focusing on.
AIM: Please tell us about the digital data centre that Cholamandalam has and what are the key objectives of this centre?
VA: We are building a top notch digital data centre, which will be a dedicated space where you can sit down and monitor the health of the business. This project is under the direct sponsorship of our Executive Director Arun Alagappan. His vision is to build a nerve centre with state-of-the-art technology and analytics to show everything about the business in real time. It will be a very extensive centre with AI-based alert generation and forecasting engine.
We are building a data lake, a Chola wide data lake, which will ingest data from HR, operations, finance and other lines of businesses. All this data will come together to allow us to create new KPIs and new performance indicators. We can mix and match operations data, to see the performance of accounts from our collection systems.
We are trying to bring all this information in one place to create an amazing facility which can have a journey of the account from start to end in front of our eyes and every touchpoint that happened to this account will be in front of us, once this project is completed.
AIM: What are the initiatives that you are working in the coming years?
VA: We are pushing the envelope on a daily basis. We are refining our goals and are in constant search of fintech partners to bring in agility to bring advancements in tech-driven decision making. We are willing to talk to more customers to bring better offerings for them.