InsTech - insurance & innovation with Matthew Grant & Robin Merttens

Automated underwriting: a pioneer’s perspective (393)

Feb 8, 2026

Guest

Marek Shafer

Marek Shafer, Vave leader in cyber underwriting; Tom Squires, AEGIS/Opal insurance executive who digitised US property products; Jonathan Spry, Envelop Risk co-founder building AI-driven underwriting. They discuss where automation actually adds edge, why complex risks often benefit most from algorithmic approaches, platform and data discipline lessons, and how generative AI and specialist underwriters reshape underwriting workflows.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

Cyber Risk Suits Machine Learning

Cyber underwriting fits AI because threats are persistent, adaptive, and human-driven.
Marek Shafer explains attackers evolve rapidly across global tech stacks, so forward-looking ML creates a predator-prey feedback loop underwriters can't achieve alone.

ANECDOTE

Opal Grew By Digitalizing A Known US Product

Tom Squires built Opal from an existing US property product and digitalized a simple quote flow.
He stresses prior product knowledge and distribution experience supercharged digital adoption across UK and US wholesalers and MGAs.

INSIGHT

Sweet Spot Is Complex Homogenous Volume

Complexity plus homogeneity plus volume is the sweet spot for algorithmic underwriting.
Marek Shafer points to high-cat US surplus lines where similar insured interests repeat across many policies but small location differences materially change risk.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

What actually makes automated and enhanced underwriting work in practice?

In this episode, three early movers in automated underwriting share hard-earned lessons from building digital underwriting propositions that have survived real market cycles. Rather than theory or hype, this conversation digs into where technology genuinely creates advantage, where it does not, and how underwriting judgement remains central even in highly algorithmic models.

Drawing on experience across cyber, US property and digital facilities, the panel explores why complexity, not commoditisation, is often where automation delivers the greatest edge. From AI-driven cyber underwriting to high-cat surplus lines property and digitally distributed specialty products, each speaker explains how they chose their focus and what they learned along the way.

Key themes include the role of data discipline in sustaining AI-led underwriting, why platform design matters more than speed to market, and how underwriters’ roles are shifting from generalists to specialists embedded in algorithmic decision making. The discussion also tackles unstructured data, submission quality and why “no data, no deal” may become a defining principle of future underwriting models.

What you’ll learn in this episode:

Why complex risks can be better suited to automated and augmented underwriting than simple, commoditised ones
How AI and machine learning are being applied in live underwriting decisions, not just analytics
The importance of volume, homogeneity and risk differentiation when building algorithmic models
Lessons from re-platforming early digital products and avoiding long-term technical debt
How generative AI is changing data cleaning, exposure management and submission handling
What enhanced underwriting means for underwriter skills, careers and decision making

Featuring perspectives from Marek Shafer of Vave, Tom Squires of AEGIS London and Jonathan Spry of Envelop Risk, moderated by Matthew Grant of InsTech.

You can also watch the video version of this panel here.

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