ResiQuant. The wordmark that property carriers now trust to read a building the way a seismic engineer would - before the ground ever shakes.
Two Stanford earthquake engineers built the software that tells insurers which structures survive a disaster - and which do not. This is ResiQuant.
Somewhere right now, a property underwriter is looking at a submission for a warehouse in earthquake country. The paperwork says concrete, three stories, built before anyone worried much about seismic code. That is roughly all the paperwork says. On the strength of it, someone is about to decide whether to insure a building worth tens of millions of dollars.
ResiQuant thinks that is a strange way to run a $200 billion market. So it built AI agents that pick up where the paperwork stops - reading inspection photos, aerial imagery and ordinary public visuals to find the structural weaknesses that decide whether a building stands or falls. The engineer's eye, at software speed, applied to every submission instead of a lucky few.
The story starts at Stanford's John A. Blume Earthquake Engineering Center, where in 2020 two PhD students kept crossing paths. Omar Issa was modeling how communities recover after disaster, borrowing machine learning to do it. Francisco Galvis was studying why certain older steel-frame buildings fail in ways their blueprints never predicted. Both had done the unglamorous work - Galvis walking the rubble after the 2023 Turkey earthquake and Hurricane Ian, Issa a second-generation structural engineer who grew up around post-disaster inspection.
They noticed the same uncomfortable thing from two directions: the people financially responsible for buildings understood them the least. Insurers priced catastrophe risk off broad hazard maps and thin submission forms. The specific building - its framing, its cladding, its quiet vulnerabilities - stayed invisible until the day it mattered most.
"Property carriers make billion-dollar decisions with dangerously incomplete data. We're transforming this by delivering engineering-grade analysis."
- Dr. Omar Issa, Co-Founder & CEOThe timing made the pitch for them. Reinsurance costs have climbed roughly 50% since 2020. Natural-disaster losses hit $217 billion in 2024. Carriers responded the way cornered businesses do - by pulling out of whole states, leaving homeowners scrambling. The industry's instinct was retreat. ResiQuant's bet was the opposite: that if you could actually see which buildings were resilient, you could keep insuring the ones that deserved it, and price the rest honestly.
That reframing matters. Resilience stops being a slogan and becomes a number - something an underwriter can act on and a building owner can be rewarded for. Reinforce your structure, pay less to insure it. It is climate adaptation smuggled in through the actuarial back door.
ResiQuant started with earthquakes - the founders' home turf - then widened into a multi-hazard platform. Today its models reason across the disasters that actually empty a reserve fund:
ResiQuant did not build a chatbot. It built specialized AI agents, each doing one job at engineering-grade depth. Carriers deploy the ones they need, and pay in proportion to their volume.
Automates the messy front door of underwriting - reading and structuring incoming submissions so backlogs disappear and processing time drops by roughly 95%.
Analyzes site inspection photos, aerial imagery and public visuals to flag the structural weaknesses and attributes that determine a building's survival.
Applies each carrier's own underwriting guidelines automatically, keeping risk selection consistent and freeing underwriters for judgment calls.
The layer underneath: computer vision and machine learning trained on real structural inspection procedures, turning imagery into building-level risk.
A second-generation structural engineer with hands-on post-disaster inspection experience. His Stanford work applied machine learning to how communities recover from catastrophe.
An expert in building resilience who conducted field assessments after the 2023 Turkey earthquake and Hurricane Ian, and studied failure modes in older steel-frame buildings.
The round funds what the founders call the boring, decisive work: extending the platform across every major US hazard and growing the engineering and AI teams behind it. LDV Capital, which invests at the intersection of visual data and AI, led on the thesis that computer vision plus structural engineering is a combination the insurance industry has never actually had.
ResiQuant's early believers are the kind of carriers that live or die by catastrophe math. Golden Bear Insurance Company - writing commercial property, earthquake and casualty across 37 states - put the platform to work on its underwriting. Skyway Underwriters partnered to rework how coastal wind gets underwritten. Enterprise trust like that is not won with a deck; it is won when the product survives contact with a real portfolio.
"ResiQuant is transforming how we process submissions, delivering risk-specific engineering insights at remarkable speed."
- Michael Brown, VP of Property, Golden Bear Insurance CompanyOmar Issa and Francisco Galvis meet at Stanford's Blume Earthquake Engineering Center.
ResiQuant founded in San Francisco, pairing structural engineering with AI.
$4M seed round announced, led by LDV Capital.
Golden Bear Insurance adopts the platform across its 37-state footprint.
Skyway Underwriters partnership targets coastal wind underwriting.
Return to the underwriter and the concrete warehouse in earthquake country. The paperwork still says almost nothing. But now the photos speak. ResiQuant's agents have already read the framing, checked the vulnerabilities against the hazard, run it through the carrier's own guidelines, and handed back a risk-specific answer before a human could finish the coffee.
The billion-dollar bet is still a bet - it always will be. The difference is that someone finally looked at the building. In an industry retreating from the places that need coverage most, that small act of seeing is the whole idea. ResiQuant did not make the risk go away. It made it legible.