Kayzad Hiramanek, Chief Operating Officer, Edelweiss Life Insurance, on why the next phase of insurance technology is not simply about automation, but about connecting data, improving decision-making and knowing where human judgement still matters.
Insurance has traditionally been a business where technology operated behind the scenes—automating processes, improving efficiency and helping organisations manage large volumes of customer and policy data. But as artificial intelligence and machine learning move deeper into the insurance value chain, technology is beginning to influence not just how processes are executed, but how decisions are made.
For Kayzad Hiramanek, Chief Operating Officer, Edelweiss Life Insurance, that shift represents a fundamental change in the role of technology leadership. Having worked across sectors including telecom, hospitality and insurance, Hiramanek says one of the most important lessons from his career has been the need for organisations to continuously adapt.
“Every business obviously goes through its ups and downs, its peaks and troughs. What is very important, especially when you’re climbing the ladder, is being adaptable,” he says.
For insurers, this adaptability is increasingly being tested by changing customer expectations, growing data volumes and the rapid emergence of AI-led decision-making.
From efficiency to transformation
The technology function in insurance was once largely associated with getting processes right, improving efficiency and ensuring compliance. According to Hiramanek, that mandate has now expanded into a transformation role, where technology is expected to create an impact across the business.
At Edelweiss Life, this transformation journey included Project Udan, which was built around three pillars: keeping the customer and convenience at the centre, improving distributor effectiveness, and making customer interactions more customised.
The technology layer supporting this transformation goes beyond individual applications. The company invested heavily in building its data infrastructure, including a data repository and data lake where information from different source systems could be correctly placed, defined and made explainable.
The resulting data environment, which Hiramanek refers to as the “Dataverse”, now supports reporting, business intelligence, risk profiling and critical business decisions.
This is an important distinction in the enterprise AI conversation. AI may be the visible component of the transformation, but the ability to use AI effectively depends on the quality, structure and governance of the data underneath it.
“Today, it’s not … what do I do with the data and the technology that I have at my disposal to make a decision to transform the organisation?” Hiramanek says.
Building intelligence into the insurance lifecycle
At Edelweiss Life, technology investments are being positioned as force multipliers rather than simply tools for process automation.
Hiramanek says 100% of the company’s policy proposals are now digital, while integrations with the broader public API ecosystem, credit bureaus, the Aadhaar stack, the Insurance Information Bureau and account aggregators help simplify onboarding through consent-based access to information. Business rules, reinsurer integrations and customer-profile data further support the process.
The result is an onboarding process in which low-value coverage can, according to Hiramanek, be issued in around 30 minutes.
But the more significant change is the movement towards using technology throughout the customer lifecycle.
The company’s MyZindagi platform, for instance, supports distributors across onboarding, training, daily routines, lead management, conversions, pitches and follow-ups. A separate need-discovery tool helps distributors understand customer needs and aspirations and tailor their conversations accordingly.
This reflects a broader shift in enterprise technology—from isolated automation to interconnected workflows where data generated at one stage of the customer journey informs decisions at another.
AI moves into underwriting
One of the areas where this becomes particularly significant is underwriting.
The core of insurance remains the assessment of financial and medical risk. AI and machine learning can help insurers process large volumes of information and identify patterns across customer and risk profiles.
Hiramanek says models are run across data collected during onboarding, while information from the wider insurance ecosystem can also contribute to the assessment. The Insurance Information Bureau, for example, provides industry-level data that can help identify customers who may have previously applied for insurance elsewhere or filed claims with other insurers.
The objective, however, is not to eliminate the underwriter.
Where the data indicates that a case requires further scrutiny, a trained underwriter can intervene and engage with the distributor or customer.
That human intervention is significant because insurance decisions can involve context that is difficult to reduce to a single data point.
The emerging model is therefore less about replacing underwriting expertise and more about giving that expertise better information, faster.
Why faster claims start with better underwriting
Claims are often regarded as the ultimate test of an insurer’s customer experience. They are also where the quality of decisions made earlier in the policy lifecycle becomes visible.
Hiramanek says Edelweiss Life recorded a 99.31% claim settlement ratio in the last financial year and had zero claims pending for settlement at the end of that financial year. He also says 84% of eligible claims that meet the eligibility criteria under the company’s Instaclaim process are settled within 24 hours.
Technology is an important part of this process, including data mining, machine learning and AI-assisted processing.
But the underlying philosophy is broader.
The insurer’s ability to settle a claim quickly is linked to the work undertaken at the beginning of the policy lifecycle—understanding the customer, assessing risk, maintaining accurate profiles and continuously managing the relationship.
In this model, claims automation is not an isolated technology project. It is the downstream result of a more connected insurance architecture.
Moving from detecting fraud to preventing it
The same principle applies to insurance fraud.
Rather than treating fraud as a problem that begins when a suspicious claim arrives, Edelweiss Life’s approach, as described by Hiramanek, involves assessing risk from the beginning of the customer lifecycle.
Multiple sources of information feed into models that generate a customer profile score. This can help determine whether the level of cover being requested is appropriate and whether the policy has been correctly matched to the customer’s circumstances and stated needs. Where the system identifies something requiring additional investigation, human intervention comes into play.
This creates a more preventive approach to fraud management.
The technology is not simply looking for suspicious transactions after the fact. It is attempting to establish a stronger understanding of the customer before the claim ever happens.
The data outside the enterprise
Yet, even as insurers become increasingly data-driven, one of the remaining challenges sits outside the insurer’s own technology environment.
Hiramanek says technology itself is no longer the primary constraint to near-instant claims settlement. The bigger limitation is information that remains outside the insurer’s ecosystem—including health records, death certificates, medical records and police records.
This highlights an important next step for digital insurance.
The ability to settle a claim instantly is not determined only by how quickly an insurer can process information. It also depends on how quickly relevant information can be accessed, with appropriate consent and governance, from the wider ecosystem.
As more public and private records become digitised and interoperable, insurers could potentially reduce some of these external dependencies.
In other words, the next challenge may not be building another AI model. It may be connecting the systems that AI needs to make a decision.
Human judgement remains part of the architecture
For all the emphasis on AI and automation, Hiramanek sees a clear role for people in the future insurance model.
He argues against the idea that AI will simply replace human beings, particularly in life insurance, where policies represent long-term contracts and commitments.
“We are AI-assisted … we are a human-driven company assisted by AI,” he says.
Routine and transactional activities can increasingly be handled by AI, according to Hiramanek, while expertise, empathy and judgement remain areas where human involvement is important.
That distinction could become increasingly important as insurers scale their use of AI.
The question for technology leaders is therefore not simply how much of the insurance lifecycle can be automated. It is where automation adds value, where AI should assist human decision-making, and where human judgement should remain central.
For insurance, the technology transformation may ultimately be less about creating a completely automated enterprise and more about creating an intelligent division of labour between data, machines and people.
The winners of that transformation will not necessarily be those that deploy the most AI. The more fundamental challenge is building the data foundations, ecosystem connections and governance needed to use AI responsibly—and ensuring that technology makes the customer’s experience simpler rather than simply making the organisation more automated.