Max Life Insurance deploys new-age underwriting and issuance system based on cloud-native architecture

​Max Life has been consistent in its resolve to integrate tech innovations in all business aspects, including underwriting. As part of its endeavor to become reliable insurers, the company has invested in building multiple AI/ML-driven underwriting and fraud management systems backed by the digitalization of forensic controls. These intelligent systems determine 360-degree UW risks and enable automated UW decisions for a substantial portfolio share. As a result, the claims paid ratio has risen steadily over the past five years, going from 97.81 percent in FY’17, to 99.34 percent in FY’22.

“In our vision to ​become a digital powerhouse, we have implemented new-age underwriting and issuance system based on cloud-native architecture. This system helps reduce human intervention and leverages AI and ecosystem partnerships to ensure instant policy issuance with superior risk management,” informs Suhail Ghai, Chief Digital & Information Officer, Max Life Insurance.

​​Minimize fraud upfront through Shield’ predictive underwriting decision system

Strengthening the sharpness of the UW risk assessment leveraging the AI/ML technologies, Max Life Insurance developed the ‘Shield’ predictive underwriting decision system, which has helped the company minimize fraud by checking customer credibility at the application stage. The Shield model leverages 10+ machine learning models to determine the likelihood of early claims and frauds at the policy application stage. The identified high-risk policies are triggered for a differentiated UW process automatically.

Rather than rejecting claims at the payout stage, the Shield model enables Max Life to proactively identify and isolate risky, fraudulent policies. This has resulted in savings of approximately Rs 800 crores in claims payout and increased customer satisfaction scores. The Shield model also helps correlate geographical locations with risk data, given how certain areas are at a higher risk of fraud. For policies from high-risk PIN code locations determined by the Shield model, the underwriting team conducts thorough risk assessments enabling stricter underwriting controls.

“To further enhance the fraud risk assessment and prevent diagnostic centre fraud, ​we ​have built a Medicheck solution, a real-time Diagnostic Centre fraud identification system to identify fraudulent practices used for undertaking customer medicals. The solution leverages machine learning models to help identify mirror reports sent by any Diagnostic center, along with assigning a statistically calculated health score for each incoming profile, explains Ghai.
​​

​Building  core AI/ML solutions​