InsTech - insurance & innovation

AI experimentation is everywhere in insurance. But how much of it is actually making it into production and delivering measurable value? 

In this episode, Robin Merttens speaks with John Hele, Co-Founder and Executive Chairman, and Greg Framke, CEO and Co-Founder of Portage AI, about why they believe the biggest opportunity for AI lies not in automating the easy work, but in tackling some of insurance’s most complicated processing and data challenges. 

John and Greg bring experience from both sides of large-scale technology transformation. Before founding Portage AI, they worked together at Resolution Life on a major technology programme to establish independent operations following acquisitions in Australia, New Zealand and the US. The transformation involved around $400 million of investment and was designed to deliver annual savings of $150 million. That experience, alongside their work applying AI in production, helped shape their view of where the technology can create meaningful value today. 

Greg explains why Portage is deliberately targeting the difficult “last 20%” of processing: work that has historically been too complicated, expensive or technically challenging to automate. Rather than starting with the technology, their approach is to identify a specific problem, train models on a client’s own data and demonstrate value through a production-ready solution before asking the client to commit. 

The conversation explores one of the insurance industry’s longest-running challenges: turning fragmented, inconsistent data into something genuinely useful. In specialty insurance and reinsurance, information can arrive from brokers, clients and other parties in different formats and using different terminology. John explains how AI can help establish a common ontology across that information, potentially overcoming data cleansing and standardisation problems that insurers have spent years trying to solve. 

But better processing is only the starting point. Once detailed underwriting and claims information is structured and accessible, John and Greg believe insurers can begin applying predictive models to identify patterns that humans may struggle to find. John uses the example of “Move 37”, the unexpected move played by AlphaGo against world champion Lee Sedol, to illustrate the possibility: could AI uncover similarly unexpected insights in insurance data that lead to better pricing, risk selection and loss ratios? 

They also discuss where this is already being applied. Greg explains the opportunity in post-bind processing, bordereaux and delegated authority business, particularly across London, Bermuda and the US. He also discusses Portage AI’s acquisition of technology and people from Insurwave and how bringing together pricing submission processing with post-bind capabilities could ultimately make operational data more useful for underwriting decisions. 

The bigger ambition is not to replace insurers’ existing technology stacks. Instead, Greg describes an AI engine that can sit between systems, receiving and structuring information before feeding it into the applications that need it. With AI developing rapidly, he also explains why remaining technology-agnostic could matter, allowing different underlying models to be selected and trained depending on the problem being solved. 

Finally, John and Greg explain why they see AI as a much bigger change than another wave of process automation. John believes the industry is entering a fourth industrial revolution and that, within the next three to five years, some insurers will begin using these technologies to grow market share profitably. Their message to insurers is simple: move beyond experimenting with AI, apply it to meaningful business problems and demand demonstrable value from the technology. 

In this episode you’ll learn: 

Why John and Greg believe AI represents one of the biggest technology shifts of their careers 

What their experience delivering a major insurance technology transformation taught them about putting AI into production 

Why some of the greatest opportunities for AI could lie in the difficult “last 20%” of insurance processing 

How AI can help extract, harmonise and structure information from complex insurance datasets 

Why creating a common ontology could unlock more value from underwriting and claims data 

How better structured operational data could ultimately contribute to pricing, risk selection and loss ratio improvements 

Why specialty insurance, reinsurance, delegated authority and bordereaux processing are particularly interesting applications for AI 

How Portage AI approaches accuracy, explainability, security and human oversight when deploying models 

Why Greg believes AI should integrate with existing insurance technology rather than requiring insurers to replace their core platforms 

What AlphaGo’s famous “Move 37” can teach insurers about the potential for AI to uncover patterns and opportunities humans may have missed 

Why John believes the next three to five years could reveal which insurers are successfully turning AI into competitive advantage 

Why insurers should move beyond proofs of concept and demand measurable business value from AI 

If you like what you’re hearing, please leave us a review on whichever platform you use or contact Robin Merttens on LinkedIn. 

Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning.

Introduction 

What happens when you build an MGA from scratch with today’s AI capabilities, rather than trying to layer them onto legacy technology? 

In this episode, Robin Merttens speaks with Theo Butt, CEO of Kinetic Insurance Services, Andy Roberts, CTO of Kinetic Insurance Services and Guillaume Bonnissent, CEO of Quotech, about Convex’s new hybrid carrier-MGA platform, the technology being built behind it and what an AI-native underwriting operation could look like in practice. 

Theo explains why Convex created Kinetic to extend into lines of business that are complementary to its existing underwriting appetite. The model is designed to give entrepreneurial underwriters some of the advantages of building their own MGA, while providing access to cornerstone capacity, technology, operations, claims and compliance infrastructure. The aim is simple: allow underwriters to spend more time building profitable portfolios and less time building the machinery around them. 

Andy takes us inside Kinetic’s technology strategy. Starting with a greenfield environment has given the team an opportunity to design around agentic AI from day one, with the ambition of automating as much non-revenue-generating work as possible. That could mean workflows running from the moment an email arrives through to quote generation, while keeping the underwriter in control of the decisions that require human judgement. 

But building an AI-native insurance business is not simply about automation. Andy explains why explainability, auditability and human oversight have to be designed into agentic workflows from the beginning, rather than added later. Every action needs to be traceable, with controls over what agents can do and clear visibility into how and why work has been completed. 

Guillaume explores what could come next. As large language models become the interface through which employees access information and complete tasks, he believes underwriters could increasingly start their day in their company’s LLM rather than moving between email, underwriting systems and other applications. The technology to do this already exists, he argues, making adoption, governance and cost control increasingly important challenges. 

The conversation also looks at one of the biggest advantages and risks of building on the technological frontier. Kinetic does not have the legacy infrastructure that can make deploying AI at scale difficult, but the pace of change means today’s new technology can quickly become tomorrow’s legacy. Building for flexibility and continuous change may therefore be just as important as choosing the right technology today. 

In this episode you’ll learn: 

Why Convex chose a hybrid carrier-MGA model for Kinetic 

How the model aims to give entrepreneurial underwriters more time to focus on underwriting 

Why Convex intends to provide cornerstone capacity to Kinetic’s underwriting cells 

How Kinetic is building agentic AI into its workflows from the outset 

Why explainability, auditability and human oversight are critical when deploying AI in a regulated market 

How automation could change the roles of underwriting and operations teams 

Why flexible technology matters when new underwriting teams need to be launched quickly 

How large language models could change the way underwriters access their systems and information 

Why AI adoption may now be a bigger challenge than technological capability 

How legacy infrastructure can make deploying agentic AI at scale more difficult 

Why the accelerating pace of technological change is forcing insurers to rethink what it means to future-proof their architecture 

If you like what you’re hearing, please leave us a review on whichever platform you use or contact Robin Merttens on LinkedIn. 

Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning.

Introduction

AI is already changing underwriting, but is the industry focusing on the part of the job where it can create the most value? 

In this episode, Matthew Grant speaks with Jamie Wilson, VP of Strategy at hyperexponential, about the evolution of underwriting technology, the move from AI-assisted tools towards agentic workflows and what 350 senior underwriters across the UK and US say they actually want from AI. 

Jamie explains why much of the value delivered by AI so far has been concentrated on efficiency and reducing manual administration. Yet when underwriters were asked what makes someone exceptional at their job, speed barely featured. Instead, they pointed to qualities such as risk selection, technical depth, judgement and commercial instinct. It raises an important question: rather than simply helping underwriters work faster, can AI help them make better decisions? 

The conversation explores what changes as underwriting moves from AI-assisted to agentic. Instead of an underwriter controlling the workflow and calling on AI for individual tasks, an agentic system can move work through the underwriting process itself, involving a human when a decision requires expertise or falls outside predefined guardrails. Jamie discusses why those guardrails will be critical and how different portfolios could require very different levels of human involvement. 

They also examine the implications for underwriting teams. As technology takes on more routine work, experienced underwriters could spend more time applying their expertise to complex risks, while technical portfolio teams may play a growing role in determining where and how greater autonomy should be used. 

Finally, Jamie shares one of the most striking findings from the research: the "succession paradox". Underwriters identified the loss of senior judgement before it can be passed to the next generation as their biggest future concern, yet coaching junior colleagues ranked last among the investment priorities explored in the survey. Could AI help bridge that gap by making accumulated underwriting knowledge and expertise more accessible to the next generation? 

In this episode you’ll learn: 

Why focusing on efficiency alone could underestimate the potential of AI in underwriting 

What underwriters believe separates exceptional underwriting from simply working faster 

The difference between AI-assisted and agentic underwriting 

How guardrails can determine when an AI agent acts autonomously and when an underwriter steps in 

Why different portfolios may require very different approaches to agentic underwriting 

How greater automation could allow underwriters to focus their expertise on more complex risks 

Why existing technology architecture and point solutions could complicate end-to-end agentic workflows 

What the "succession paradox" reveals about the industry's approach to developing junior underwriting talent 

How AI could potentially help transfer expertise from experienced underwriters to the next generation 

Book recommendation:  

Thinking, Fast and Slow by Daniel Kahneman for its insights into human judgement, data-driven decision-making and the psychology of adopting new technology. 

Read hyperexponential's Underwriting Edge Report

Discover what 350 senior commercial P&C underwriters say defines underwriting excellence, where AI is delivering value and what could hold the industry back in hyperexponential’s Underwriting Edge 2026 report.

Curious for more? Catch-up with all that happened at hx Live NYC on 17 September.

If you like what you’re hearing, please leave us a review on whichever platform you use or contact Matthew Grant on LinkedIn. 

Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning.

Introduction 

Facultative reinsurance can give insurers the additional capacity they need to write risks they might otherwise decline. But for smaller, more transactional placements, its value increasingly depends on how quickly that capacity can be accessed. 

Joining Matthew Grant on the podcast, Dom Sindall, Head of Digital Broking, Facultative at Marsh Re, and Stephen Sheehan, Head of Analytics, Facultative at Marsh Re, explain how digital trading, automation and better data could reduce facultative placement times from days to hours or even minutes. 

Dom explains why much of the delay in facultative reinsurance comes not from the work itself, but from the elapsed time between people completing each stage of a placement. Removing some of those dependencies could make facultative reinsurance significantly more accessible for smaller risks, where speed can matter more than price. 

The discussion explores how advances in AI are helping automate another source of friction: insurance data. Stephen explains how extracting policy information from documents has become increasingly straightforward, while exposure data remains more challenging. Schedules of values (SOVs) can contain dozens of tabs, multiple currencies and inconsistent terminology, making ingestion, standardisation and quality control critical parts of the automation process. 

They also examine where humans should remain involved. Fully automated workflows offer greater speed, but Dom and Stephen explain why firms need to determine what level of data quality is ‘good enough’, where validation gates are required and which capabilities should be built internally or bought from technology partners. 

Looking ahead, Dom explains why the growth of digital trading could expand the facultative market itself. Reinsurers that invested early in data standards have been able to move quickly, while the next wave may increasingly look to partnerships to accelerate their own digital capabilities. 

In this episode you’ll learn: 

Why speed is becoming increasingly important for smaller facultative placements 

How automation could reduce placement times from days to hours or minutes 

Why AI has made extracting policy data significantly easier 

Why exposure data remains a more difficult automation challenge 

How to decide where human validation belongs within automated workflows 

Why ‘good enough’ data can be more commercially valuable than pursuing perfection 

How insurers and brokers can approach build versus buy technology decisions 

Why early investment in data standards is creating an advantage in digital trading 

How reducing friction could help grow the facultative reinsurance market 

Book recommendations: 

Stephen Sheehan: Don’t Believe Everything You Think by Joseph Nguyen 

Dom Sindall: The Innovator’s Dilemma by Clayton Christensen 

If you like what you’re hearing, please leave us a review on whichever platform you use or contact Matthew Grant on LinkedIn. 

Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning.

Insurers are moving quickly from asking what AI can do to asking how they can actually make it work in underwriting. But experimenting with AI is one thing. Running it reliably across an insurance business is another. 

In this episode, Matthew Grant speaks with Andy Moss, GM Underwriting at Duck Creek and former Co-founder and CEO of Send, following Duck Creek’s acquisition of the underwriting technology business. 

Andy explains why successful enterprise AI needs to run “on rails”, supported by the infrastructure, governance and workflows required to move beyond proofs of concept and into production. He explores how underwriting workbenches are evolving into orchestration engines and why that shift could become increasingly important as insurers begin introducing AI agents into their workflows. 

The conversation also examines a growing challenge for the industry: how to preserve underwriting knowledge as experienced professionals retire. With underwriting becoming more complex, Andy argues that insurers need to codify the processes, referrals and institutional knowledge that often still exist in people’s heads, creating a technology foundation that can support both underwriters and AI. 

They also discuss how insurers should approach the increasingly important question of whether to build or buy technology. Andy’s advice is simple: “Buy the commodity and build the differentiators.” Rather than recreating established systems, insurers have an opportunity to focus their development efforts on the data, processes and underwriting expertise that genuinely give them an advantage. 

Finally, Andy shares the story behind Send’s acquisition by Duck Creek, how a search for growth investment became an acquisition in around 120 days and what building and selling Send taught him about the importance of assembling a team that genuinely cares about the problems customers are trying to solve. 

In this episode you’ll learn: 

Why enterprise AI needs to run “on rails” to succeed in underwriting 

How underwriting workbenches are evolving into orchestration engines 

Why AI proofs of concept can struggle to make the transition into production 

How agentic AI could change underwriting workflows and automate previously manual tasks 

Why capturing institutional knowledge is becoming increasingly urgent as experienced underwriters retire 

How technology can codify the processes and knowledge that currently sit inside underwriting teams 

Why insurers should “buy the commodity and build the differentiators” 

Where internal technology teams can create the greatest competitive advantage 

What bringing underwriting orchestration and core insurance systems together could enable 

How Send’s search for growth funding developed into its acquisition by Duck Creek 

Why Andy believes shared values and genuine commitment to customers were critical to building Send 

Andy’s recommendations: 

Invest Like the Best with Patrick O’Shaughnessy 

Desert Island Discs 

Read InsTech's recent report with Send on 'Underwriting orchestration engines: equipping insurers for the AI age' here.

Register for InsTech's upcoming 7 October Evening Event with Send and Duck Creek here: Bridging between the islands of automation.

If you like what you’re hearing, please leave us a review on whichever platform you use or contact Matthew Grant on LinkedIn. 

Sign up to the InsTech newsletter for a fresh view on the world every Wednesday morning.