Global MosaicResolution 1 km²
The spatial data
your agents
rely on.
Global Mosaic indexes every square kilometer on Earth across fifteen harmonized data layers — queryable by shape, boundary, or coordinate. Built API-first so AI agents can ground their reasoning in the real, physical world.
Pixel-level intelligence,
built for machines.
Every 1km² cell on Earth is pre-indexed, pre-aggregated, and queryable in milliseconds — ready for the agents, pipelines, and workflows that drive modern industry.
Define a shape
or boundary.
Pass any geometry — a coordinate pair, polygon, ZIP code, county, state, or country. The API resolves to every 1km² cell that intersects your shape.
POST /v1/query
{
"geometry": "POLYGON(...)",
"layers": ["population",
"income",
"land_use"]
}Receive aggregated
spatial data.
Within seconds, get statistical summaries across every cell in your boundary — mean, median, percentile distributions, trend direction, and raw cell-level values.
// response
{
"cells": 847,
"pop_median": 4821,
"income_trend": "+3.2%/yr",
"dominant_use": "residential"
}Agent grounds
its decision.
Your agent receives rich, structured spatial context and acts on it immediately — selects sites, flags risk, generates reports, triggers downstream workflows. Grounded, not hallucinated.
// agent decision
target_grids: [
"47.2°N, 122.4°W",
"47.3°N, 122.5°W"
]
reason: "income +4.1%/yr"
"low density"Every layer that matters,
at 1 km resolution.
Each layer is harmonized, gap-filled, and indexed to the same 1km grid. Any combination can be queried together in a single call — agents receive structured machine-readable data, not image tiles.
What agents can do with spatial intelligence.
Designed for large-volume, programmatic usage across industries where location context changes the quality of every decision.
Detect redevelopment before the census sees it.
Identify 1km² grids with rising land values, increasing median income, and growth in prime working-age populations — a systematic edge over market consensus.
Pinpoint the highest-growth grids in seconds.
Favorable income-to-rent ratios, strong density, demographic demand for new residential or mixed-use development — ranked candidates in a single query.
Find undervalued pockets next to premium areas.
Detect adjacent grids with similar demographics but significantly different land values. Agents surface these discrepancies automatically across whole portfolios.
Overlay climate, income, and value in one call.
Identify high-value assets that carry hidden environmental or socioeconomic risk — informing acquisition due diligence and portfolio stress testing.
Simulate upzoning before the public hearing.
Model how increasing density in specific 1km² grids would affect population capacity, land value, rent levels, and displacement risk.
Evaluate proposed uses at granular scale.
Check whether new commercial, industrial, or residential development aligns with surrounding demographics, land values, and community needs.
Locate underutilized land across a city.
Grids where land value is high but building density is low — combined with income trends to flag the highest-leverage sites for rezoning.
Protect vulnerable neighborhoods proactively.
Identify low-income or historic areas where downzoning may help preserve affordability before market pressure triggers displacement.
Refine premiums below the ZIP.
Assess property value, climate exposure, and income resilience at 1km² scale. Each cell carries independent risk signals that aggregate into accurate pricing.
Price localized risk at pixel resolution.
Grids with similar ZIP codes but different elevation, income stability, or flood exposure can carry meaningfully different premiums.
Map portfolio exposure cell by cell.
Understand geographic concentration risk down to individual grids — informing reinsurance, capacity limits, and risk transfer grounded in sub-ZIP reality.
Monitor climate drift across the book.
Continuously track physical risk layer changes across every policy. Automatically flag grids whose risk profile has materially shifted.
Track urban form at 1km, over time.
Population and land use changes at granularity finer than administrative boundaries — sprawl, infill, and decline patterns invisible at ZIP level.
Reveal income cliffs and opportunity deserts.
Fine-grained inequality measurement across contiguous grids — segregation gradients standard area-based methods smooth over.
Map who carries disproportionate risk.
Overlay climate, elevation, income, and demographic layers to surface grids where vulnerable populations bear environmental burden.
Compare countries without harmonizing data.
Because every layer shares the same global 1km grid, urban patterns and income distributions compare directly across borders.
The spatial layer your
agents reach for.
Global Mosaic is a first-class tool in any AI agent's toolkit. Via MCP or REST, agents query spatial boundaries and receive structured, aggregated data in seconds — every decision grounded in the real physical world.
Agent receives task
“Find the top 5 ZIPs in the Southeast US with highest income growth and low building density — mixed-use candidates.”
Calls Global Mosaic
Agent issues a structured MCP query across Southeast US boundaries, requesting income trend, building density, and population growth layers.
Receives spatial context
In ~1.5 seconds the API returns ranked, grid-level data — median income YoY, density per km², population trend by cohort.
Delivers grounded output
A ranked site list with full spatial justification. Not hallucinated. Every recommendation traceable to a real-world cell.
// MCP tool call from agent context
tool: "global_mosaic.query",
input: {
boundary_type: "zip_code",
region: "southeast_us",
layers: ["income_trend", "building_density", "population_growth_25_40"],
sort_by: "income_trend_desc",
filter: { building_density: "<0.4" },
limit: 5
}
// response arrives in ~1.4s
results: [
{ zip: "30309", income_trend: "+5.8%/yr", density: 0.28 },
{ zip: "37203", income_trend: "+4.9%/yr", density: 0.31 },
{ zip: "28202", income_trend: "+4.4%/yr", density: 0.27 },
// ...
]