This article is by Amir Massoudi, Chief Product Officer at Cover Genius
Amir Massoudi, Chief Product Officer at Cover Genius comments: The next breakthrough specialty product won’t come from a traditional underwriting room. It will be spotted by an AI agent monitoring millions of transactions, and built before the market even knew there was a gap.
Traditional MGA product cycles take 12 to 18 months. Agentic AI is compressing that dramatically, but only for platforms that have the data to power it. Embedded insurtechs sit inside the transaction flow. At Cover Genius, AI systems process demand signals across hundreds of millions of purchase events in 60+ countries in real time, before a broker has raised a hand or an underwriter has run a feasibility study. Most MGAs simply don’t have that capability.
The next generation of specialty products won’t be designed in an underwriting room; they’ll emerge from behavioural patterns that only embedded platforms can see.
We’re excited about an agentic AI capability that moves beyond optimization into net-new product creation. One agent segments customers into distinct personas from partner data; another tests which benefits resonate with each cohort and packages them accordingly.
Our partnership with a leading independent digital rail and coach technology platform is an early proof point with a bespoke, co-created cancellation product designed around real customer travel behaviour. It is precisely the kind of product that emerges when embedded data, not assumptions, drives the brief.
That said, speed without accuracy is dangerous and our confidence layer is what makes this scalable: Cover Genius’ platform lets partners test new products across the full customer journey, from purchase through to claims, verifying both commercial appetite and genuine value before anything goes live.
On the claims side, we already use agentic AI with a human-in-the-loop model, processing at speed while maintaining compliance guardrails. The reality is the AI is ready. What’s held the industry back isn’t the technology, it’s the distribution infrastructure to test a novel product at real scale. That’s the differentiator traditional MGAs can’t replicate quickly
Be the first to comment