SATQUERY/ AI

Ask the Earth.Trace the evidence.

Evidence-backed satellite image analysis.

Local control plane

Backend capability status

This readout comes from the live health service and permitted model registry.

Connecting to localhost
Mission selection

Choose how to
read the scene.

The imagery and the question define the route. SatQuery selects the correct observation mode without exposing model plumbing.

OBS-01
Single scene

Observe

“What features are visible in this scene?”

Scene description + grounded answer
CMP-02
Two dates

Compare

“Where has the built-up area changed?”

Change mask + measured regions
FUS-03
Optical + SAR

Fuse

“Map water and vegetation in this area.”

Class polygons + confidence
Ground track / 05 Stages

From raw scene
to verified answer.

A five-stage execution path validates the input, selects a specialist, and preserves a trace of every decision.

01

Upload imagery

Add one scene, a temporal pair, or aligned optical and SAR inputs.

TRACE ENABLED
02

Ask a question

Describe what you need to understand in plain, operational language.

TRACE ENABLED
03

Interpret the mission

SatQuery validates metadata and identifies the permitted analysis route.

TRACE ENABLED
04

Run the specialist

The relevant workflow processes the scene while preserving coordinates.

TRACE ENABLED
05

Inspect the evidence

Review the answer, masks, measurements, warnings, and execution trace.

TRACE ENABLED
ANALYSIS / SQ-260901CHANGE DETECTION
ILLUSTRATIVE SCENE
NO BACKEND ARTIFACT
N
Detected changeRegion boundary
Evidence before language

The answer is only the beginning.

Inspect the geometry, measurements, confidence, and execution trace behind the generated response.

Changed area12.48 km²
Regions detected02
Specialist confidence0.91
  • Geographic masks and polygons
  • Separate specialist confidence
  • Warnings and abstention policy
  • Reproducible execution trace
Evidence-backed claims onlySensor-aware model routingReproducible execution audit