Vertesia Platform Drives Agentic AI Revolution in Insurance

As an industry that has long mastered the art of pricing future risk for individuals and corporations, the insurance sector is now faced with its own risk-reward conundrum. The promise of Agentic AI – intelligent systems that don’t just follow scripts but reason, adapt and solve problems – is on the table. Yet most carriers are leaving it there, untouched.

As Vertesia’s Chief Revenue Officer, Chris McLaughlin states, today’s market is fractured into “the haves and have-nots.” A select few AI projects are in production, while other insurers, he says, are “stuck in experimentation,” with pilots running that never scale into meaningful business value.

“For the vast majority of the market, they've got one or two apps done, sometimes with third parties, sometimes themselves,” Chris notes. “A lot of them are stuck in experimentation.”

With industry research consistently showing single-digit adoption rates for production AI applications, the underlying question becomes why insurance has struggled where other sectors have succeeded.

The usual reasons – escalating development costs, system complexity, security requirements, skills shortages – fail to account for why only 30% of enterprise organisations are in production. 

The real problem, according to Keith Schlosser – a 36-year veteran and Insurance Transformation Advisor at Vertesia – is hiding in plain sight: the crushing weight of the industry’s own history. 

Decades of mergers and acquisitions, Keith argues, have created a tangled web of legacy platforms, trapping data in silos and making innovation feel nearly impossible.

Insurers are facing a clear choice: break through the complexity or fall further behind.

The automation paradox in insurance

For decades, insurers have tried to innovate their way out of complexity one piece at a time. They buy a point solution for claims, another for renewals and a third for underwriting. Each tool promises to reduce manual work and, in isolation, it often does. But this approach has created a monster.

This is the “automation paradox” – each solution adds complexity, increasing system fragmentation and creating more – not less – work. This cycle of buying, integrating and maintaining separate tools is why conventional automation has failed to truly transform insurance.

Traditional automation is rigid. Operating through predetermined logic paths, these systems are a series of ‘if-then’ statements that can fail the moment a situation veers off-script. If a claim amount exceeds a set threshold, it gets routed to an adjuster. If a policyholder’s address is missing, it gets flagged.

Agentic AI represents a fundamentally different approach. It doesn’t follow a script; it follows a goal. Chris explains, “agentic tools excel in complex processes working with a variety of different documents where we’re asking the AI agent to reason its way through a specific task.” It’s designed to handle the unpredictable reality of insurance, where no two cases are ever truly the same.

Keith believes agentic AI isn’t just another technological advancement but a transformational advantage for insurers, and sees it as the key to finally breaking free from the industry's fragmented past. “I see it really as a foundational technology that allows for the end-to-end joining together of all these processes, which in large part have been done independently of each other,” he states.

Platform economics versus point solutions

The economic argument for a platform approach requires analysis of why insurance companies have consistently chosen point solutions despite their integration costs. An AI platform capable of building AI agents requires a comprehensive infrastructure that enables autonomous agents to perform substantive work across multiple business functions rather than addressing isolated use cases.

“What we’re finding is that getting an LLM to reason is not necessarily difficult,” Chris explains. “What is critically important is giving it the right tools with which to perform a task.”

This observation suggests that LLMs' reasoning capabilities have become commoditised, shifting the competitive advantage to implementation infrastructure.

The platform approach addresses an obsolete procurement model: “Gone are the days where you have to buy that one bespoke solution or piece of software that handles a finite number of use cases,” says Keith.

However, the persistence of this procurement pattern suggests institutional resistance that extends beyond technology evaluation to organisational behaviour and risk management frameworks.

Claims processing as a proxy for industry transformation

Claims processing provides a useful lens for analysing agentic AI’s potential impact because it is one of the most complex insurance processes – combining regulatory requirements, documentation analysis, internal and external inputs and time-sensitive decision-making.

Keith offers the example of demand letter responses as one potential area for significant efficiency gains in the end-to-end claims process. Traditional processes require substantial time for documentation collection, coverage analysis and response formulation.

“With the Vertesia platform, we can decrease the amount of effort to respond to a demand letter down from days to a couple of hours,” Keith reports. 

However, the significance extends beyond time compression to the nature of the work being automated.

The value proposition in the claims process encompasses customer satisfaction and retention through faster claim resolution, as Chris notes: “The faster the claim settles, the more efficiently it settles. Ideally, we never want to get to the point where we have to process a demand letter.”

“Organisations that can process claims more quickly reduce claims payouts and the cost of processing claims while, at the same time, delivering a superior customer experience.”

This creates a multiplier effect where operational efficiency improvements generate revenue protection benefits and a far more efficient end-to-end process for the carrier.

Claims reserving represents another application where automation could address systemic industry challenges. Keith observes that current processes are “incredibly labour-intensive and often flawed.” 

The combination of labour intensity and accuracy problems suggests opportunities for technology solutions that improve both efficiency and quality – a convergence that justifies investment.

“By using the Vertesia platform, our clients can dramatically improve reserving accuracy, fine-tune reserves as needed and allow skilled workers to focus on the most valuable tasks while reducing the potential for reserve charges that negatively impact earnings,” Keith explains.

Extension to underwriting and back-office transformation

The extension to underwriting processes reveals the benefits of a platform like Vertesia's. 

While there are some products that enable straight-through processing for routine applications, only the autonomous AI agents built with the Vertesia platform expand automation capabilities to complex scenarios without requiring manual intervention. 

This breakthrough allows for automation of high-touch, complex underwriting, similar to advancements in low-touch, high-flow underwriting.

“The staff will almost immediately understand how to leverage the tool in the context of their process, challenges and opportunities for improvement,” Keith says.

The breadth of use case possibilities suggests that the platform approach could address a systemic issue in insurance technology investment, where there has been a tendency to focus resources on customer-facing and revenue-generating functions while under-investing in operational infrastructure.

Client operations, compliance reporting and finance functions often operate with constrained resources despite their importance to business performance. 

The economic logic becomes compelling when considering the cumulative impact across multiple functions. 

The platform approach enables organisations to address operational challenges using consistent methodologies and shared infrastructure, generating returns that exceed the sum of individual improvements.

Implementation and adoption dynamics

A “workshop-first” implementation approach addresses a critical challenge in enterprise AI adoption: demonstrating value quickly enough to sustain organisational commitment. 

“The goal of a workshop is to help customers identify high-value, low-effort use cases, and then show them how quick and easy it is to build AI apps and agents that have an immediate impact on their organisation. These solutions can typically be deployed in less than a week,” Keith says.

This rapid deployment capability addresses what Keith identifies as a persistent enterprise challenge: “I cannot tell you the number of times I've been in a meeting where the business wants to do something, but the challenge on the IT side is how much effort it will take to accomplish their business goal.”

Vertesia can unlock creativity and break down the barriers that many carriers have had to work around due to constraints that they've learned to just accept.

However, production deployment requires addressing accuracy, security and operational control requirements that distinguish enterprise environments from sandbox settings. 

Vertesia has implemented multiple safeguards, including enabling AI agents to ask humans when they need more information rather than just making up an answer.

Human-in-the-loop is a critical design consideration. “The value proposition is to enable employees to work on things that are of higher value to the organisation and to their customers,” Chris explains.

This framing of autonomous AI agents as augmentation to humans rather than replacement may be essential for adoption in an industry where decision-making accountability carries regulatory and fiduciary implications. 

Another productive approach is to employ agents to validate the work of other agents, increasing the accuracy of outputs without requiring human intervention.

ROI assessment and measurement challenges

Traditional ROI frameworks may inadequately capture value creation that spans efficiency improvements, quality enhancements and strategic capabilities. The challenge lies in quantifying benefits that accrue across multiple business functions and time horizons.

A platform approach generates compounding returns as organisations deploy multiple applications, but measuring these network effects requires analytical frameworks that extend beyond conventional project ROI calculations.

Time-to-value emerges as a critical factor in this analysis, though specific timelines vary based on implementation complexity and organisational requirements. 

The ability to demonstrate value quickly becomes essential for sustaining investment and organisational commitment through implementation challenges.

Industry adoption and competitive dynamics

The executive-level normalisation of AI discussions that Keith describes suggests that adoption decisions increasingly focus on implementation approach rather than technology adoption itself.

Customer acceptance represents a particular consideration given the insurance industry's trust-dependent business model. 

The need for human oversight in complex decisions is critical. “There will always be a place for a human to validate what an autonomous agent is doing,” says Keith.

This requirement for human authority in critical decisions may shape technology design and implementation approaches.

The accelerated investment in AI solutions draws parallels to the cloud computing evolution, suggesting that competitive advantage may accrue to early adopters who develop organisational capabilities while technology costs remain high for custom implementations.

“Where we see the real opportunity is for organisations to have a very standardised, programmatic approach to building and running these applications,” Chris explains.

Strategic implications and market evolution

Keith's anticipation of a broad range of use cases suggests that initial implementations may influence a broader organisational transformation. 

The economic enabling effect on previously unfunded improvement opportunities could create competitive differentiation for organisations that develop systematic approaches to identifying and implementing AI opportunities.

Keith identifies particular potential in back-office functions: “Insurance companies have a lot of processes in the back office that oftentimes are not funded but can offer tremendous value to the organisation.”

The convergence of technical capability with strategic necessity creates conditions where competitive pressures may accelerate adoption beyond normal technology diffusion patterns.

Chris concludes: “We really think we're just scratching the surface as far as generative AI goes.”

The analysis suggests that agentic AI's impact on insurance may extend beyond operational efficiency to enable new business models, competitive strategies and industry structures. 

The successful implementation of agentic AI platforms like Vertesia's provides organisations with standardised infrastructure and a unified approach for deploying AI solutions while eliminating silos, enhancing governance and reducing maintenance and costs.

To read the full article in the magazine, click HERE.