New Delhi, Nov. 18 -- For health insurers, a claim is often the first significant interaction with a customer after a policy is purchased. Aditya Birla Health Insurance is attempting to change that relationship by using data and artificial intelligence (AI) to identify health risks well before they translate into hospitalisation or claims.

"A claim is usually the end of a health journey that began months or even years earlier," Mayank Bathwal, chief executive officer of Aditya Birla Health Insurance tells Mint CIO Circle in an exclusive interview. The company combines health assessments, wearable data, diagnostics and claims information, with customer consent, to identify early indicators of conditions such as diabetes, hypertension and high cholesterol.

Its predictive models classify customers by risk and trigger interventions ranging from wellness incentives and digital nudges to personalised coaching and clinical support. The insurer has generated more than 28 lakh WellBeing Scores, reached over 2.7 lakh high-risk lives and estimates that its interventions have helped prevent approximately 12,000 hospitalisation days.

The approach reflects a broader shift in its business model, from being primarily an insurer that pays for hospitalisation to becoming a health partner that engages with customers throughout their health journey. The business grew its gross written premium by 39% year-on-year to Rs 6,855 crore in FY26, according to company results.

Building a data-led health ecosystem At the centre of the strategy is a cloud-agnostic, modular technology architecture organised across five layers. These comprise core insurance systems for policies, underwriting, billing and claims; an integration layer connecting hospitals and external health services; a data layer bringing together health, behavioural, engagement and claims information; an intelligence layer powering risk stratification, fraud detection and claims automation; and customer-facing platforms led by the Activ Health app.

The integration layer connects the insurer to approximately 17,900 hospitals, diagnostic centres, pharmacies, wearable platforms and the National Health Claims Exchange (NHCX).

Having started operations in 2016 as a digital-first insurer, Aditya Birla Health does not carry the legacy technology constraints associated with some older insurance businesses. However, it continues to renew its core systems while building differentiated capabilities around health engagement and risk intelligence.

"Core systems like policy administration, underwriting and claims need reliability, security and cost efficiency, so we modernise them on a modular, API-led architecture," Bathwal says. "The goal is a stable core and a fast-moving engagement layer." This approach also helped the company launch Activ Yuva.

The insurer builds capabilities that differentiate its health-first model, including the Activ Health app and its HealthReturns engine, while partnering with external providers for services such as diagnostics, teleconsultations, wearables, pharmacies and fitness. Its rewards model also draws on the approach of joint venture partner Momentum Group, adapted for Indian customers. It is leveraging shared infrastructure such as NHCX and Bima Sugam to extend integration and reach.

From data to preventive intervention The company's predictive intelligence operates across the customer lifecycle, not just healthcare. Propensity models and AI-enabled voice calls help identify customers at risk of lapsing, while machine learning models and generative AI-generated case summaries support faster underwriting assessments.

On the health side, one data platform tracks customers across three risk categories: healthy, borderline and high-risk. The interventions vary accordingly. Healthy customers earn HealthReturns through programmes such as Active Dayz and the Healthy Heart Score, with rewards of up to 100% of the premium on select plans. Around 9,500 policyholders earned back their entire premium last year.

For borderline customers, models identify indicators such as high body mass index (BMI) or declining activity, prompting personalised nudges, goals and coaching. High-risk customers receive more human-led support through the Chronic Management Programme, with technology equipping care managers to monitor progress and support recovery.

The insurer completed more than 6.6 lakh health assessments and intervened with over 2.7 lakh lives during the year. Among coached Chronic Management participants who submitted two laboratory reports, 85% improved or maintained their biometrics. The company also feeds these outcomes back into its models to improve subsequent decisions. "Data only matters if it leads to timely action," Bathwal says.

The Activ Health app, built in-house on a cloud-agnostic, microservices-based architecture, is central to this engagement. It has crossed 6.1 million downloads, with customers spending more than 74 minutes a month on the platform. APIs connect it to policy information, claims, health records, teleconsultations and care managers, while data from wearables, fitness apps and partner gyms feeds into its activity and rewards programmes.

"Given the sensitivity of health information, the company says the data is encrypted and customers retain control over it," says Bathwal.

In FY26, the insurer engaged 13 lakh retail customers through personalised journeys. Half of its app users accessed four or more tools, while retention improved 30% over FY25.

Technology is also being used to address barriers to preventive healthcare. Customers can book free health assessments and consult doctors through the multilingual app. Activ Yuva extends outpatient department (OPD) cover to consultations and diagnostics, while its preventive programmes use measures such as the Healthy Heart Score and WellBeing Score to make health information more actionable. HealthReturns also incentivises healthier behaviour, with Activ Yuva extending the proposition to nutrition and sleep.

Automating claims and detecting fraud Beyond prevention, the insurer is using AI and analytics to improve claims efficiency and identify suspicious activity. With more than 24 lakh claims handled in FY25, automation is important to meeting regulatory timelines for cashless authorisation of one hour and discharge approval of three hours.

Customers can submit and track claims through the app, website and WhatsApp. NHCX integration reduces manual paperwork, while AI extracts information from documents and generative AI produces structured summaries for claims adjudicators. AI-assisted adjudication now covers 62% of in-house cashless claims, while approximately 72% of the company's claims are cashless.

Fraud detection uses a layered approach, beginning with non-disclosure models at underwriting and extending to red-flag rules and machine learning models at pre-authorisation. Network analytics identify patterns across hospitals, doctors and policyholders, while generative AI helps investigators summarise discharge, laboratory and radiology documents. Bathwal says these initiatives have generated savings of more than Rs 300 crore. The company estimates that nearly 15 out of every 100 claims may be non-payable because of fraud, non-disclosure or abuse.

The objective, he says, is to direct scrutiny towards suspicious cases without making the claims experience more difficult for genuine policyholders. "Automation gives us scale, and human expertise adds the personal touch," Bathwal says.

For Aditya Birla Health, the larger ambition is to make the insurer-customer relationship continuous rather than transactional, with technology supporting healthier behaviour, better risk assessment and more efficient insurance operations well before a claim arises.

Published by HT Digital Content Services with permission from MINT CIO CIRCLE.