Pankaj Gupta, CTO, Bharti AXA Life Insurance, believes technology modernization has helped the company provide superior customer experience by improving the customer experience at every touchpoint – reducing customer onboarding TAT, simplifying the customer journey by minimizing the number of clicks, improving customer verification process which is simpler and intuitive, improving medical scheduling with automation for hassle free medical process and more.When it comes to analytics, it is being utilised across the value chain and functions at Bharti AXA Life, right from understanding the customer, customer acquisition, customer engagement, retention and win back, bringing in operational efficiency and improving distribution effectiveness with an objective to improve topline, cost optimization, profitability and customer experience.
ML models have been used to develop a personalized strategy to maximise the objective function. Given the limited resources to tackle any problem, the output of such models helps the organization in targeting efforts to areas/ customers where maximum value could be derived.
Big data analytics is being used in end to end persistency suite from onboarding to deep lapse revival for streamlining our efforts to improve persistency/renewal collection throughout the policy journey.
It is also being used in early claims model to predict the likelihood of an early claim and risk based rejection at the underwriting stage for every policy.
“While ML based predictive models are not a novel concept, the following differentiators would make our solutions unique: Combination of multiple predictive models to yield a composite score; Augmentation of model scores with industry data; Model building rigor and incorporation of behavioral data. It uses 100+ variables over a 5- year time frame for training and testing; Model accuracy – ability to identify cohorts as small as 5% of total issuances contributing to 50% of all early claims; Real time deployment with an automated workflow for differential underwritings such as auto-reject for very high risk, mandatory physical verification for medium risk and STP for no risk cases,” said Gupta.
Why Implement ML?
Bharti AXA Life decided to invest in ML models in August 2019 and set up a dedicated Analytics and Business Intelligence function. Some of the initial requirements which propelled the organisation to think about using Analytics and ML models were related to optimization of existing resources for identification of policies at risk of early claims and drive persistency.
“Further, we had a homogenous strategy towards solving such business problems, e.g., investigation of majority of cases pre-issuance or use a more heuristics based approach to pick and choose policies to investigate, homogenous communication strategy across customer segments, collection strategy irrespective of the customer’s likelihood to renew. Thus, leading to wastage of precious company resources due to not having a holistic view of different factors which impacted a business problem,” added Gupta.
Key benefits gained
By leveraging ML, the company has been able to witness: Reduction in early claims, hence positively impacting the profitability; Reduction in physical investigations by concentrating on genuinely risky cases, leading to significant cost savings; Improvement in renewal collections, increasing the overall revenue; Decreased policy issuance TAT, leading to enhanced customer experience and NPS; end-to-end integration including deployment on AWS cloud for real time integration with the onboarding platform and real time decision making through an automated workflow.
