Your Geospatial Agentinside QGIS.
An autonomous geospatial agent embedded inside QGIS. Describe your spatial analysis in plain English — it plans the workflow, executes the code, and delivers publication-quality maps. No GIS commands, no Python, no friction.
A geospatial agent that thinks like a GIS analyst
GeoEdge AI is a ReAct geospatial agent — it reasons about your data, acts with real GIS tools, observes the result, and iterates until the analysis is correct. Not a chatbot. A fully autonomous spatial agent.
Type what you need in plain English — no GIS jargon, no operation menus, no Python required. "Buffer the waterways by 50m and find all parcels within the buffer."
Before acting, the agent inspects every layer — field names, CRS, geometry type, sample values, feature count. It understands what you have before deciding how to proceed.
The agent selects the right tools, writes and executes PyQGIS code, runs Processing algorithms, and manages your layers — one verified step at a time.
The agent inspects your output data, chooses appropriate symbology, applies classification and labels, and renders a print-ready A4 layout — intelligently designed from the data itself.
Built for Australian spatial professionals
From environmental assessments to site suitability — GeoEdge AI handles the workflows that take hours and makes them conversational.
Clip vegetation datasets to project boundaries, classify plant community types, calculate clearing extents, and produce regulatory-grade maps for development applications.
Run multi-criteria suitability analyses, overlay zoning, flood, bushfire, and infrastructure constraints, and produce scored suitability maps for planning reports.
Chain any sequence of buffer, clip, intersect, dissolve, join, and raster operations. The agent handles CRS mismatches, null geometry, and edge cases automatically.
Process satellite imagery, calculate NDVI change, overlay with cadastral and tenure data, and generate monitoring reports with trend maps — repeatable and audit-ready.
Describe the map you need and GeoEdge designs it — choosing the right classification method, colour ramps, label placement, and layout for your data and purpose.
Access and process publicly available satellite data from Sentinel-2, Landsat, and MODIS — calculate spectral indices like NDVI, NDWI, and NBR, clip to your project boundary, and export analysis-ready rasters without leaving your workflow.
Every action is observed.
Every result is verified.
GeoEdge AI doesn't fire and forget. Each tool call returns structured observations — feature counts, attribute summaries, geometry statistics, error messages. The agent reads them and decides whether to continue, retry, or adjust its approach.
This means complex multi-step workflows succeed even when the data has edge cases — mismatched projections, null values, geometry errors. The agent handles them automatically.
REASON — Form a plan
Agent analyses your request and inspects available layer schemas before deciding which tools to use.
ACT — Execute one tool
One atomic action at a time — processing algorithm, PyQGIS code, layer management, symbology.
OBSERVE — Read the result
Structured observation returned: feature count, field values, CRS, extent, any warnings.
ITERATE — Refine if needed
Agent decides: correct and continue, or backtrack and try a better approach.
Everything a GIS analyst needs
From data inspection to print-ready export — GeoEdge AI covers the full spatial analysis workflow without leaving QGIS.
Agent reads field schemas, geometry types, CRS, sample values, and feature statistics before every analysis — no more "wrong field name" errors.
Access to all QGIS Processing providers — native, GDAL, GRASS, SAGA. If it exists in Processing, GeoEdge AI can run it.
Your spatial data never leaves your machine. All analysis and processing runs locally within QGIS — the agent only exchanges task instructions with the cloud, not your datasets.
For operations beyond standard tools, the agent writes and executes validated PyQGIS code in a sandboxed namespace — safely and transparently.
Data-aware map design — single, categorised, graduated, and rule-based symbology chosen to match your data type and purpose. Not just a default style.
Full A4/A3 layouts with north arrow, scalebar, legend, and attribution — generated automatically at the end of every analysis. Export as PNG or PDF.
Every action logged with a JSON audit trail — undo any step, replay a workflow, export for team review. Non-destructive by design.
The agent learns from its history. Successful approaches, field naming conventions, and local QGIS quirks are retained — so it gets smarter over time.
See token consumption per session, estimated cost before each run, and real-time usage against your plan quota — no surprise bills.
Built for Australian regulatory workflows
GeoEdge AI is the only AI-powered QGIS tool with Australian regulatory datasets pre-loaded and natively understood by the agent — not just searchable, but reasoned over.
Simple, token-based pricing
Pay for what you use. Every plan includes the full QGIS plugin with transparent cost estimates before each run — no surprise charges.
Loading plans...
Start your first analysis in minutes
Install the QGIS plugin, connect your free account, and describe your first spatial workflow. No code, no menus, no friction.