In 2024, natural catastrophes cost the global economy $417 billion, according to Gallagher Re. Behind that number sits a quieter problem: most large companies cannot actually prove their own risk. They know it is there. They pay to cover it. But when it matters most, they often discover they got it wrong.
We invested in ScyAI because we believe that gap is about to close, and that ScyAI is building the platform to close it.
The thesis:
Insurance runs on trust, and trust needs evidence
Every large company that owns physical assets, factories, power plants, warehouses, data centers, relies on three things being true at once. The insurance policy actually pays when a loss hits. The price paid reflects the real risk, not an insurer's rough average. And the policy satisfies the covenants, the conditions, in the company's loan agreements with its banks.
Usually, at least one of those three is wrong. And companies tend to find out the hard way: a claim that pays out half of what was expected or a lender that calls a breach because an insurance clause quietly fell out of compliance at renewal.
The reason is structural, not a failure of the people involved. A company's asset data, its insurance policies, and its credit agreements live in different systems, written in different languages, reconciled by hand maybe once a year. The risk teams doing this work are typically small and overstretched, armed with little more than spreadsheets. Meanwhile, insurers price on averages, because building a model specific to one company's assets has simply been too expensive to justify.
Our conviction is that this work is about to be transformed. Not by replacing the risk team, but by giving it the analytical firepower it has never had.
What ScyAI is building
ScyAI is best understood not as a climate tool, but as a risk intelligence platform: an AI teammate that sits a layer above the world's climate and catastrophe models. It takes their outputs and applies them to what a specific company actually owns, insures, and owes.
The platform is agentic, meaning it doesn't just answer questions on command, but plans and carries out multi-step analysis on its own. A risk manager describes what they need in plain language ("assess the flood exposure across this portfolio"), and ScyAI plans the work, pulls the right data, and runs the right tools, from geospatial analytics to flood simulation.
What it delivers are three concrete outcomes that the insurance market actually accepts:
- Coverage that pays. ScyAI reads every clause of a policy against a company's real exposure and against court-tested wording, surfacing the hidden gaps that turn into denied claims.
- The fair price. Not simply a cheaper premium, but the right amount of cover at the technical price of the risk. The positioning is precise: same risk, better data, lower premium.
- Covenants that hold. Every insurance policy is checked against the company's loan agreements, so that a routine renewal never accidentally triggers a costly breach with a lender. This is the outcome that is least optional of all.
Crucially, every step ScyAI takes is logged and explainable. In insurance, that matters more than anything. An underwriter will not act on a black-box number, so ScyAI doesn't produce one. The risk manager keeps the pen. What changes is that they now command the equivalent of an eight-role expert bench, from policy-wording counsel to catastrophe modeller, without adding a single new hire.
Why this matters:
From cost center to provable advantage
Consider a mid-sized industrial company with plants across several countries and tens of millions in insurable value. Today, its risk manager assembles a renewal submission by hand, largely trusting the broker's and insurer's view of the company's own risk. With ScyAI, that same manager walks into the renewal with an auditable, data-backed case: this is our real exposure, this is where our current coverage falls short, this is the technical price of our risk.
The relationship shifts. The company stops being a passive price-taker and starts negotiating from evidence. Over time, that points toward a market where insurers compete to underwrite risks that have already been clearly quantified and documented, rather than pricing blind and padding the number for uncertainty.
Why this team
What convinced us most is that this team has already lived the problem it is solving.
Founder Bernhard Rannegger previously built and scaled Swiss Re's enterprise physical-risk platforms, including a landmark joint venture with Palantir Technologies. He is joined by risk expert Alex Sidorenko, named Risk Manager of the Year by FERMA (the European federation representing thousands of risk professionals), whose proprietary quantitative methodology has a long track record of unlocking multi-million-dollar premium efficiencies and better coverage terms for global multinationals, earning corporate risk awards from RIMS (the global risk management society) along the way.
That last point captures the opportunity precisely. The methodology is proven; Sidorenko has delivered exactly this kind of analysis by hand, worth millions a year to the companies he worked with. What has been missing is the technology to scale it. Backed by an engineering team drawn from Apple and other technology leaders, ScyAI is turning a proven, elite risk methodology into a scalable enterprise software category.
Why we're excited
ScyAI sits exactly where PT1 Ventures likes to invest: at the point where a critical part of the real-world economy still runs on manual work and rough averages, and where the right technology can replace guesswork with evidence. The category is early and execution will decide who leads it. But the combination of a proven methodology, deep founder-market fit, and a genuinely differentiated agentic platform is exactly the kind of bet we like to make.
Find out more about Scy AI over at www.scyai.com.