How P&C insurers create continuous transformation from AI assistants to production impact

Most insurers see the potential: faster underwriting, smarter claims triage, dynamic risk modeling, and new levels of customer responsiveness. The challenge lies in execution — how to modernize core systems, integrate emerging technologies, and deliver continuous improvement without introducing risk or disruption to the business.

The shift to engine-first architecture

An engine-first approach is a mindset and an architectural strategy that prioritizes the underlying capabilities — the “engines” power policy, billing, claims, and data — over the interfaces that sit above them.

In an engine-first model, insurers evolve incrementally, upgrading functional layers while keeping the business running. The approach emphasizes interoperability, composability, and continuous delivery — allowing carriers to innovate without the “rip and replace” cycle that has historically accompanied core modernization.

Engine-first design also sets the stage for AI to thrive. When data flows freely and processes are decoupled, insurers can embed intelligence directly into workflows, whether that’s automating document ingestion, augmenting underwriting with machine learning insights, or optimizing claims with predictive analytics.

Faster updates: Redefining agility

Traditional IT models in insurance are defined by release cycles — large, infrequent, and often disruptive. As a result, business teams are perpetually waiting for the next update while operations grind through testing and deployment.

A model delivering smaller, more frequent, and largely invisible updates to end users is preferable. Insurers can roll out updates over a weekend without business interruption and empowers insurers to move at the speed of their customers, responding to market changes, regulatory shifts, or new distribution opportunities in real time.

Several carriers have demonstrated what this looks like in practice: thousands of system enhancements deployed in months, hundreds of configuration changes released in days, and large-scale transformations delivered without downtime.

Normalization of data: The hidden accelerator

Legacy systems, mergers, and siloed operations have left many carriers with fragmented data ecosystems, where underwriting, claims, and distribution each operate on separate truths.

Normalizing data across the enterprise is critical to scale AI adoption. When carriers achieve real-time, standardized data models, they can train models on consistent information, embed insights across workflows, and monitor outcomes with confidence.
A normalized data layer becomes the connective tissue for decision-making, enabling a single, transparent view of risk, performance, and opportunity — moving AI from experimentation to everyday operation.

Transformation at scale: Balancing speed and stability

Many insurers are proving it’s possible to modernize rapidly while maintaining absolute stability. Insurers can perform zero-disruption updates, introduce thousands of new capabilities in months, or enable teams to deploy hundreds of business rule changes within days. Transformation at scale doesn’t have to come at the expense of reliability. When designed correctly, modern architectures make insurers more resilient, not less.

The pragmatic automation roadmap

As AI maturity grows, the industry must also take a pragmatic view of automation. Full autonomy is neither realistic nor desirable in most insurance functions. The near-term opportunity lies in agent-assisted intelligence — where machines handle repetitive or data-heavy tasks, and humans provide oversight, empathy, and governance.

Analysts estimate that roughly 80% of underwriting tasks could become AI-augmented within five years. Yet human judgment will remain central, especially where regulatory compliance, liability, or ethical considerations are involved. The goal is not to replace expertise but to amplify it. Automation should free human talent to focus on higher-value activities: complex risk assessment, customer relationships, and innovation. The balance between human and machine will define the next generation of high-performing carriers.

From projects to platforms

The insurers best positioned for the future will operate more like technology companies — delivering constant, incremental improvement rather than episodic change. Engine-first architecture, version-less delivery, and real-time data normalization will help insurers sustain transformation, integrate AI responsibly, and scale innovation across the enterprise. The road to AI ROI is about building a foundation where innovation becomes routine, change becomes safe, and technology evolves continuously alongside the business it supports.

Opinions shared in this piece are the author's own.

Nick Lien is Head of Product at Duck Creek Technologies, where he leads strategy and development for core solutions serving the global property and casualty insurance market. He has more than 20+ years in software & innovation—helping industries like insurance modernize and adapt to evolving technology, meeting carriers wherever they are on their transformation journey.

See also:

• How AI and location intelligence drive more equitable home insurance
• Legacy tech is starving innovation in the insurance industry
• Over a third of Americans fear AI bots have final say on insurance claims