For a company that describes itself as a creature of habit, HDFC Life has a habit worth noting: about every ten years, it resets its technology architecture. Not because anything is broken, but to stay relevant. The latest reset, Project Inspire, is now the foundation for one of the more ambitious enterprise AI programs in Indian insurance.
Sameer Yogishwar, the company’s Chief Operating Officer, sits at the center of it. He has been with the HDFC group since 1998 and with HDFC Life since 2000, including nearly a decade in Dubai as founding CEO of a subsidiary. He became COO in 2025 and now oversees operations and technology, as well as strategic alliances, corporate communications and international business. That combined remit shapes how the company builds. “When there is one person who’s owning both the process and in some sense also the platform, we have a smooth operating process,” he says. “We don’t have a business versus IT kind of boundary.”
The results were on show at InsurAI 3.0, HDFC Life’s flagship AI event, themed “Human AI Collaboration.” The company and its partners presented 33 enterprise-level innovations across sales, service, underwriting, claims and operations, grouped under nine technology themes: Voice AI, Vision AI, Developer AI, Agentic AI, Sales AI, Tech Innovation, Complaint Analytics & Customer Service, Do-It-Yourself and Employee Productivity. According to Yogishwar, 20 of the 33 are GenAI use cases are already deployed with measurable outcomes.
From Inspire to Intelligence
Launched in January 2023, Project Inspire moves HDFC Life from a policy-centric, monolithic model to a modular, customer-centric one, built on cloud-native architecture and digital-first delivery. It runs across nine transformation tracks, each with its own owner, covering areas such as product configuration, customer communication, CRM and data platforms, and headless customer engagement, with a final track integrating them.
At the foundation are modular platforms, heavy API integration with partners and a unified data platform that Yogishwar calls “a single source of truth” for everything entering, originating in and leaving the company. In FY26 the payoff included zero-touch policy issuance, same-day claim settlement for credit protect products, faster underwriting and a shift toward real-time incentive and commission management. The next targets are instant policy issuance and same-day payouts for both claims and commissions.
Yogishwar is clear that the models are not the hard part. “About 70 to 80% of any AI effort that an insurer does is working towards data quality and integration, not on the models itself,” he says. HDFC Life avoids rip-and-replace modernization. It wraps its core systems in APIs, decouples the data and then modernizes progressively. “Legacy becomes a constraint only if you allow legacy to start dictating your architecture,” he says.
Proving value, not just launching pilots
Yogishwar’s yardstick is outcomes. “Launching all these fancy things without being able to look at value in terms of the outcomes that we’ve delivered is frankly an exercise in futility,” he says. Drawing on the company’s investor presentations, he cites 16% higher frontline productivity in partner-facing areas powered by AI in the first quarter, and about 60 lakh (six million) customer conversations handled entirely by AI last year. In policy servicing, decision-making is 30% faster, and on the support side, email click rates are up 52%.
Every roadmap item, technology or process, also faces a single test: does it remove friction for the customer or the partner? If not, “it just waits.” He measures that friction against far more than other insurers. Customers who breeze through an airport with Digi Yatra or a mobile banking app carry those expectations into buying a policy. “The competitive pressure on the journey side is not coming from other insurance companies,” he says.
The same logic applies to HDFC Life’s 500-plus distribution partners across banks, NBFCs, broking firms and individual agents. The goal is a standardized API layer, configurable journeys instead of custom builds and onboarding measured in hours and days rather than months. “They all have different front ends, but they all hit the same engine,” Yogishwar says.
The AI credit: reinvesting time instead of cutting teams
The most distinctive part of HDFC Life’s approach is what it does with the productivity gains. Yogishwar bristles at a pitch he hears often from vendors: improve productivity by 20%, then cut 20% of your people. “That’s not the way we look at it,” he says.
Instead, he talks about an “AI credit.” Every efficiency gain hands a team hours back, and the strategic question is where to spend them. “You get that one extra hour… How do you use that AI credit to now impact some other area and do things slightly better?”
Medical underwriting offers the clearest example. An AI model now reads more than 50% of ECG reports, a share that continues to grow, and issues a decision without each report having to pass through a doctor. The rollout was deliberately cautious. “Not even one incorrect case should go through,” Yogishwar says, and the team kept testing until accuracy met that bar. The company’s reinsurer is also comfortable with the approach. Doctors are not being written out of the process. Their time is being redirected to other parts of the medical workflow that need improvement.
Other deployments follow the same pattern of AI assisting and people deciding:
Complaint prediction: identifying what could become a complaint and where communication falls short, so customers get the right information up front. Yogishwar notes the industry’s image problem. “Nobody speaks about the 99.8% of claims that we have settled immediately, but everybody speaks about the 0.2%.”
Fraud detection: analyzing early claim trends, pin codes and profiles linked to organized fraud, which he likens to “a chess game.” A wrongful payout, he says, eats into the genuine customer cohort.
Assisted underwriting: AI support across both financial and medical underwriting to speed up decisions.
Voice AI: GenAI voice bots supporting customer service, cross-sell and persistency.
These capabilities are built by HDFC Life’s own teams and scaled through a unified governance framework. GALAXY 2.0, the company’s platform for governed AI adoption, provides the common foundation.
An insurer run by agents, guided by people
Looking ahead, Yogishwar has few doubts. “Agentic AI will be the top priority. There’s no two ways about it,” he says. The key word, in his view, is discipline. “Today’s AI assists a person. An agentic AI completes the task.”
His picture of HDFC Life in 2030 has agents running onboarding, servicing and claims journeys end to end, with humans handling judgment, empathy and exceptions. He expects agentic workflows, a real-time data fabric and hyper-personalized engagement to define the industry.
He does not expect the winner to be whoever has the most AI. “We will have not just a lot of AI, but we will also have data, we will have the governance, and we will have the operating model, which is actually ready to operate in an environment of overall trust,” Yogishwar says.