New Delhi, Aug. 19 -- Bandhan Group-owned private life insurer Bandhan Life Insurance is expanding artificial intelligence (AI) beyond underwriting into policy persistency, complaints, claims and agent productivity, as it prepares to scale its business, according to chief technology officer (CTO) Sumanta Ghosh. The insurer, formerly Aegon Life Insurance, has built a cloud-native architecture on Amazon Web Services after discontinuing new business on its legacy platform in 2019. Its microservices-based, event-driven technology stack supports round-the-clock operations without batch-processing downtime and enables more than 100 deployments a month. The architecture has also helped speed up policy issuance. Around 80% of non-medical, low- and medium-ticket applications are issued on the same day, while customers receive digital policy documents within two to four minutes of approval, according to Ghosh. The technology push comes as Bandhan Life expands under its new ownership. The insurer reported a 107% increase in new business premium to around Rs.500 crore in FY26 and is targeting about 30% growth in FY27, according to the company. It has also widened its portfolio beyond term insurance to include participating, non-participating and unit-linked products. In an exclusive interview with Sohini Bagchi, Ghosh discussed the insurer's technology architecture, the maturity of its AI-led underwriting and the digital initiatives planned over the next two years. Edited excerpts: Bandhan Life inherited an existing insurance business. How much of its technology architecture has been rebuilt? Since 2019, we have not written new business on the legacy platform. We built a new cloud-native, microservices-oriented and event-driven architecture on AWS. Unlike conventional life insurance systems, we have no batch-run downtime and can provide uninterrupted operations. We undertake more than 100 continuous deployments every month. The omnichannel architecture allows us to launch products and services across channels and onboard partners quickly, irrespective of their technology maturity. Its event-driven design provides near-real-time updates to our self-service portal and partner systems. Connected workflows and a unified data lake also give us a strong foundation for AI adoption. Where is AI currently producing measurable results and which AI application has progressed beyond the experimental stage? AI-led underwriting is our most mature implementation. We have used data and AI in underwriting for several years. We are now at an early stage in using AI to assess the likelihood of policy persistency. We have also launched an AI-based agent productivity tool, which is currently being tested. How is technology helping agents and partners work more effectively while retaining the human touch in life insurance? Customer onboarding will continue to require human interaction because life insurance remains predominantly a push product in India. We are not looking to replace that interaction. Instead, we are using AI and other technologies to equip the sales team with better information and help it engage customers more effectively.
Where will the next phase of technology investment be directed?
We have reached an advanced stage across cloud, core platforms, data architecture, APIs, automation and cybersecurity. Our focus now is on automating less-frequent operational scenarios and embedding AI into business processes where it can create measurable value. This will help us manage scale without increasing operational complexity at the same rate.
How is Bandhan Life using its customer and policy data?
We have created a unified customer-data layer on top of our data lake. We are developing propensity models across different parts of the business cycle. Our underwriting is already mature. The next areas include predicting persistency and improving complaints management. Over time, the same data foundation can support more personalised interventions across the customer lifecycle.
With AI entering underwriting and risk assessment, where is Bandhan Life comfortable allowing AI to make autonomous underwriting decisions?
We are building AI to make underwriting faster, more consistent, and more data-driven, while retaining accountability to human underwriters. We are comfortable with autonomous decisions for data-rich and relatively low-risk customer segments, where outcomes can be validated and governed. Human oversight remains important for complex cases, higher-risk profiles and decisions requiring contextual judgement.
Which technology metrics matter most to the company?
Policy issuance time and digital conversion rates are among our most important metrics. Around 80% of non-medical, low- to medium-ticket applications are issued on the same day. Once a policy is approved, customers receive the digital document within two to four minutes. Our core policy administration, accounting and commission-generation systems also operate almost in real time.
What digital initiatives have you planned over the next 12-24 months?
Over the next three to six months, we plan to introduce a new pre-issuance video-verification process. It is being developed on a newer technology stack and designed to work smoothly on low-bandwidth networks, which is important for our customer base. We are also advancing AI use cases across persistency, complaints and claims, while automating operational edge cases. These initiatives should help us generate economies of scale as the business grows.
Published by HT Digital Content Services with permission from TechCircle.