INSURANCE
Property Risk Analysis
Property Risk Analysis. What it takes to run it.
Aerial and street-level assessment of roof condition and hazards.
// 01 · The question
“What condition is this roof in, and what is next to it?”
// 02 · Why it is hard
Imagery age varies wildly by area. A property assessed on three-year-old aerial capture is assessed on a roof that may have been replaced twice, and nothing in the image says how old it is.
// 03 · What the labels have to be
Polygon
Decided before the first frame
Underwriting wants footprint, roof area, material and proximity to hazards. Those are region measurements on an aerial view, and the polygon is also what the downstream GIS expects, so it survives the handoff unconverted.
This is the decision that is expensive to reverse. The geometry has to match what the answer contains, and finding out it does not means labelling the set a second time.
// 04 · How we run it
Three parts of one loop, on this job.
LexAnnotate
Damage Segmentation
Annotate vehicle damage, property defects, and workplace safety conditions with precision bounding boxes and segmentation. Human-in-the-loop verification reduces blind spots in complex edge cases.
LexInsight
Risk Profiling
Link visual damage assessments to claims data, vehicle databases, and property records. Track damage progression over time for workers' comp and commercial property monitoring.
LexAlert
Accuracy Alerts
Monitor deployed claims models as vehicle fleets evolve. Detect accuracy drift when new vehicle types, lighting conditions, or inspection regions change. Auto-retrain before the drift reaches the claims desk.
// 05 · What breaks it after launch
Leaf-on and leaf-off capture are different worlds for a property model. Summer canopy hides roof area and overhang risk that winter capture shows plainly, so the same house scores differently by capture date.
It is not the only one that can get this use case, it is the one that usually gets it first. All ten conditions.
// 06 · In insurance
Claims Automation Stability
The Challenge
Damage detection models struggled as vehicle types, lighting, and edge cases increased. Inconsistent visuals caused false positives.
Our Execution
Created production-ready visual datasets. Applied human-in-the-loop validation to reduce blind spots.
The Result
Stabilized accuracy across varied real-world conditions. Reduced manual claim reviews.
6,000+
Cases resolved per month