Skip to main content

Documentation Index

Fetch the complete documentation index at: https://www.mireye.ai/llms.txt

Use this file to discover all available pages before exploring further.

Insurance underwriters need more than a risk score — they need a per-field audit trail they can hand to a regulator. Mireye Earth returns flood zone status, wildfire indicators, building data, and soil risk signals sourced directly from USGS, NOAA, USDA, and USFS, with a source_url and fetched_at timestamp on every value. When a state regulator asks where a figure came from, you click the link.

​
Key fields for underwriting

FieldWhat it answers
within_floodplain_polygonIs the property inside a 100-year floodplain polygon? (USGS NHDPlus HR)
coast_distance_mHow far from the nearest shoreline? Determines storm-surge exposure. (NOAA CUSP)
surface_water_permanence_pctWhat percentage of the time is surface water present at the property edge? (JRC GSW)
FieldWhat it answers
slope_degreesWhat is the terrain slope? Steeper slope accelerates fire spread and increases runoff risk. (USGS 3DEP)
tree_canopy_pctWhat fraction of the area is covered by tree canopy? High canopy = high fuel load. (USFS NLCD TCC)
ndvi_currentCurrent vegetation greenness index — a proxy for live fuel moisture. (Sentinel-2)
ndvi_change_5yHow has vegetation density changed over 5 years? Dieback signals increased dry-fuel accumulation. (Sentinel-2)
FieldWhat it answers
primary_buildingBuilding height, footprint area, and FEMA occupancy class for replacement-cost modeling. (Overture + FEMA USA Structures)
FieldWhat it answers
bedrock_depth_cmHow deep to bedrock? Shallow bedrock limits foundation options; a proxy for earthquake vulnerability. (USDA SSURGO / Pelletier)
soil_drainage_classNRCS hydrologic group — how quickly does the soil drain? Poor drainage amplifies flood exposure. (USDA SSURGO)
soil_k_factorSoil erodibility factor. High K-factor indicates erosion and subsidence risk. (USDA SSURGO)
FieldWhat it answers
nearest_bridgeNearest bridge with NBI condition rating and year built — critical for claim-event access planning. (FHWA NBI)
Every field’s source_url points to the originating federal agency — USGS, NOAA, USDA, or USFS — so auditors can independently verify any value without contacting Mireye.

​
Wildfire underwriting with a preset

The wildfire_underwrite preset returns the six fields most relevant to wildfire exposure in a single call: lcms_class, tree_canopy_pct, ndvi_current, ndvi_change_5y, slope_degrees, and elevation. 
curl -s https://mireye-earth.fly.dev/v1/fetch \
  -H 'content-type: application/json' \
  -d '{"lat": 34.0522, "lng": -118.2437, "preset": "wildfire_underwrite"}' | jq
The response includes per-field provenance for all six values. tree_canopy_pct from USFS NLCD TCC and ndvi_current from Sentinel-2 together give you a live-fuel proxy; ndvi_change_5y flags recent dieback that dries the fuel load further; slope_degrees from USGS 3DEP tells you how fast a fire would travel uphill.

​
Flood risk with a preset

The flood_risk preset returns elevation, coast distance, floodplain status, wetland intersect, surface water permanence, and the nearest named waterbody in one call. 
curl -s https://mireye-earth.fly.dev/v1/fetch \
  -H 'content-type: application/json' \
  -d '{"lat": 29.7604, "lng": -95.3698, "preset": "flood_risk"}' | jq
within_floodplain_polygon is sourced from USGS NHDPlus HR. The source_url in the response links directly to the USGS dataset page so you can confirm methodology independently.

​
Natural-language underwriting query

Use /v1/ask to get a cited narrative answer suitable for pasting directly into a loss memo or underwriting note. The planner selects the right fields automatically based on the question. 
curl -s https://mireye-earth.fly.dev/v1/ask \
  -H 'content-type: application/json' \
  -d '{
    "lat": 29.7604,
    "lng": -95.3698,
    "question": "What is the flood and wildfire risk at this property?"
  }' | jq
The response includes an answer string with all values cited inline, and a citations array with source, source_url, and fetched_at for every field used.

​
Bulk processing

For batches of addresses, geocode each to a lat/lng coordinate — the US Census Geocoder is free and works at scale — then call /v1/fetch in parallel for each coordinate. Mireye caches responses server-side, so repeated queries for the same coordinate (common when re-scoring the same portfolio) return quickly. Use an async HTTP client or a thread pool to maximize throughput across the batch.