In the Lloyd’s market, for every £1 of premium, 47 cents is lost to attritional losses. US commercial lines pay out roughly the same — 50c/$1 of premium in everyday, non-catastrophe losses. Death by a thousand cuts, on both sides of the Atlantic.

And it’s because the way that underwriters are viewing these risks and pricing these risks is different than what it looks like today. You get a document, an excel sheet for loss history, and maybe a couple of images from the broker but who took them and when? Then you'll go and use Google Maps or Street View - which is outdated by months, sometimes years. Underwriting data is a snapshot; risk is a movie.

Phyll launched in Jan 2026 to solve this exact problem, by building a system of record for our built world off of footage the world is already capturing. Billions of cameras already pass every insured asset daily — dashcams, fleet cameras, drones, CCTV, traffic cams, phones, GoPros, smart glasses. Dashcams alone capture 750x all of Earth's roads every single week (30B miles). Phyll converts that existing footage into underwriting-grade risk intelligence. Our pipeline resolves assets across sources and over time, enabling us to provide each asset a persistent ID and track it over time, each pass building on top of the last, creating a historical archive that cannot be retroactively built. We don't just detect risk; we remember it. By watching the lifecycle of our built environment, we can then predict what will happen based on similar characteristics, attributes, and more. In the 6mo since our launch, we've been quite busy:

• Lloyd's Lab Cohort 16 participant and "Market Choice" award winner on Demo Day (07/15)
• Winner of the inaugural Risk Mitigation Startup Award at the Lloyd’s × The Hartford × Google Cloud innovation showcase (July 2026), judged by Dawn Miller (CEO Lloyd's Americas) and Mo Tooker (President of The Hartford)
• 2 paid US pilots live on commercial property portfolios — a utility-scale solar site and an 11 site industrial portfolio in North Carolina. Underwriters confirmed our data "materially impacts the way they view/price the risk", had "more confidence in the pricing", and would have remediation requirements attached to the policy due to our data. 
•  $396K in signed contracts, including a competitively awarded Innovate UK grant in collaboration with BT, Univ. of Nottingham, and Autonome (MoD company), a utility customer (512CMG), and 11 municipalities on 4 continents.
• Successful local pilot with one of the largest ride share/delivery companies in the world who now intend to "expand nationally"
• Partnerships with fleets and 2 of the top 10 fleet dashcam companies, giving us access to 100K vehicles covering 75% of the major roads in the US, EU, and Japan every single month.
• Partnerships with aerial/satellite companies, giving us updated visuals of the entire planet weekly.
• Partnership with ESRI to distribute our data to utility, insurance, and government customers of theirs
• Ongoing discussions with C-Suite at LexisNexis around a distribution deal,  partnership, and collaboration

Our team of 5 spans more than a decade of computer vision and spatial intelligence, with prior systems deployed by Fortune 500, Red Cross/FEMA, large telcos, Block, and the US DoD and UK MoD, alongside CPCU-credentialed insurance expertise across underwriting and analytics. We've delivered visuals and intelligence from the ground and the sky with teams like Snapchat and Bluesky, the leading UK aerial surveying company. Our advisory board includes Alexander Tsetsenekos, former Head of Insurance at Tesla and Luminar with 25+ years in underwriting and insurance data platforms; Sean Gorman, serial geospatial founder with 3 exits to Esri, MAXAR, and Snap; Louis Rosenberg, inventor of AR and haptic technology, with 300+ patents; Cameron Chell, CEO of Draganfly; and Paul Parkinson, the Head of Spatial at Oracle.

Stop pricing the past. Price the present.