Videos
AI Adoption Accelerates Across the Insurance Enterprise
Insights
- Insurance firms are accelerating AI adoption to improve productivity, reduce costs, and shorten delivery timelines.
- AI-enabled delivery models are reshaping software development through paired programming, testing automation, and intelligent augmentation.
- The future operating model will rely on a growing mix of human expertise and AI agents, supported by large-scale workforce reskilling.
Narahari S, Service Offering Head of Insurance at Infosys discusses how AI adoption is rapidly accelerating across the insurance industry as organizations focus on measurable business outcomes such as productivity, savings, and faster time to market. He explains how insurers are embedding AI across both run and change operations, with rising expectations for 40–60% productivity improvements over the coming years. The conversation highlights the growing role of AI-assisted development, paired programming, and intelligent automation in modern delivery models. Narahari also emphasizes that workforce transformation is becoming a strategic priority, with large-scale reskilling initiatives designed to prepare talent for AI-native operations. Looking ahead, he describes how delivery economics will increasingly evolve around a blended model of humans and agents working together to drive efficiency and scale.
AI adoption is accelerating across insurance
Narahari S:
From a responsibility standpoint, I am a service offering head. Insurance is one of the key segments under me. As far as this AI topic is concerned, I see that there's a huge adoption in many of our clients.
Productivity, savings, speed define measurable outcomes
Narahari S:
What is critical is the ability to adopt them and deliver the outcomes both on productivity, savings and reduce the time to market from the business standpoint. Many of the insurance clients are also adopting significantly and we see that ramp up.
AI talent scarcity demands strategic workforce reinvention
The talent becomes critical part of the entire thing that we are doing. We are having a big program called Talent Ascent. Talent Ascent is making sure every member goes through a complete full stack training including AI. This is around 600 to 650 hours of actual course that they take.
40-60% productivity is baseline
Narahari S:
Run and change forms a bedrock of Infosys as we all know. On the run side, the expectations from the clients are reduce the cost as much as you can. They don't want to spend almost anything on run. Expectations are 40, 50, 60% of productivity improvement over a period of four to five years. That is table stakes today.
On the development side, we are looking at paired programming, where we are using AI to help maybe a JL4, a junior developer or maybe a mid to senior developer to see how AI can help augment, finish the code, do the testing and then make him more or make him or her more productive.
Humans and agents redefine delivery economics
Narahari S:
In all our proposals and deals that we do we see a combination of human plus agents factored in. We are seeing it in many of the deals where we see the humans to start with, agents will be very less because we need to understand the ecosystem of our clients. And as we proceed over a period of a year or two, we see humans reduce, the FTE count reduce and then number of agents go up. As we see more and more of these kind of deals fructify, we will be able to deliver a lot more productivity to our clients.
