ZestyAI today announced a new partnership with Kin Insurance, a direct-to-consumer, digital home insurance provider committed to serving catastrophe-exposed markets.
As part of its expansion into California, Kin has deployed Z-FIRE, ZestyAI’s wildfire risk model, to accurately assess property-level wildfire exposure and expand access to coverage in high-risk areas.
The model was integrated in just 10 days, enabling Kin to move quickly in meeting the needs of California’s evolving market.
California continues to face a widening insurance gap, driven by more frequent natural disasters and insurer withdrawals. This partnership supports Kin’s commitment to bringing insurance options back to communities impacted by limited availability of coverage.
Z-FIRE uses machine learning to evaluate each property’s unique characteristics, including defensible space, building materials, topography, vegetation, and more, to predict which properties are most likely to experience a wildfire and which ones will survive.
“California requires a smarter, more modern approach to home insurance risk management,” said Michael McCright, vice president, pricing and risk management at Kin.
“ZestyAI’s Z-FIRE model brings the level of insight we need to confidently assess risk and offer coverage in areas at risk of wildfire.”
“Kin’s expansion into California is exactly the kind of forward-thinking move our technology was built to support,” said Attila Toth, Founder and CEO of ZestyAI. “Together, we’re helping show that with the right data and tools, insurers can re-enter challenging markets with confidence—and deliver meaningful protection to homeowners who need it most.”
“Equally important, Z-FIRE empowers insurers to recognize and reward mitigation efforts at the property level, encouraging homeowners to take action that reduces risk for themselves and their communities.”
Z-FIRE’s performance has been validated through real-world events: Following the recent Los Angeles wildfires, Z-FIRE’s highest-risk ratings closely corresponded with the hardest-hit areas, with 94% and 87% of the areas affected by the Palisades and Eaton fires rated as High Risk by the model.
Among homes located within these perimeters, those flagged as having the highest vulnerability to damage were 50% more likely to be destroyed—highlighting the critical value of property-specific insights, even in dense urban settings.
ZestyAI works closely with regulators to ensure transparency, validation, and continuous monitoring of its AI-driven models. The company has secured regulatory approvals across all Western states for Z-FIRE, along with broad acceptance of its severe convective storm suite of models in Texas, Colorado, the Midwest, and the Great Plains.