OrionAI

OrionAI

AI grounded in your data — your catalogue, forms and dashboards — so it understands your workflow and your business, not just questions. It runs the work and verifies the result before you act on it.

OrionAI summarising health facilities across Peru on a hexagon grid, with its explanation, colour key and data sources
Benefits

AI that knows your business.

Grounded in your data

One AI on your own catalogue, forms and dashboards — it understands your workflows instead of starting from a blank chat window.

Verified before it publishes

Every AI result is independently checked before anyone acts on it, so the whole team can rely on the answer.

A force multiplier, not a replacement

It removes the busywork; your specialists' judgement stays the standard its output is held to.

Features

One AI layer, everywhere the work happens.

Drafts the form

Describe a survey in a sentence and OrionAI drafts the structure, ready for you to refine in OrionForms and take offline.

See OrionForms
Runs the analysis

Ask a spatial question in plain language and the analysis is assembled against catalogue layers and returned as a map you can publish.

See Spatial Analytics
Builds the dashboard

Turn a question into a live dashboard on catalogue data — and keep asking that dashboard questions afterwards.

See OrionInsights
Verification is the default

Nothing publishes unchecked. Results are verified against the same governed layers they were built from — which is what makes them usable in a report.

How governance works
Use cases

What changes when the first draft already exists.

Arid rangeland seen from the air
A community conservancy fields a ranger patrol form without waiting on a data team

Ranger patrols record wildlife sightings, poaching and human–wildlife conflict daily — often across conservancies with no connectivity at all, and the form has to fit the patrol before the first one goes out. Describe the patrol in a sentence and OrionAI drafts the form; rangers refine it, collect offline on Android, and the submissions land in the catalogue the moment they sync.

Summer 2024 crop vigour around Fresno, California, made by OrionAI
An agronomist checks crop vigour across California's Central Valley in a minute

"Crop vigour in California's Central Valley around Fresno in summer 2024" is the whole request. OrionAI found 22 cloud-screened Sentinel-2 scenes from June to August, computed NDVI, classified it and styled it, and the result card explains what each colour means — so the agronomist starts from a finished, explained map instead of a download queue.

OrionAI mapping surface water around Tonle Sap from Sentinel-2 imagery, explaining what the colours show
A nutrition programme turns this month's screening forms into a live dashboard

The programme already collects monthly screening data in OrionForms; what it lacks is the picture on the wall when the coordination meeting starts. OrionAI builds a live dashboard on its own submissions and the catalogue's population and boundary layers — districts falling behind come back as charts and maps, verified before they reach the donor report, and the team keeps asking the dashboard questions between meetings.

FAQ

Questions, answered.

What is OrionAI?

The AI layer across OrionGIS. It drafts forms, runs analyses and builds dashboards, and every result is independently verified before it publishes — so nothing reaches your team unchecked.

Does the AI replace our GIS team?

No. It is a force multiplier: it drafts, runs and builds so specialists stop doing busywork, while their judgement stays the standard its output is checked against.

How is AI usage metered?

By AI questions per month: 10 on Free, 250 on Growth, 2,500 on Professional, and custom limits on Enterprise. Everything else on the platform is unaffected by the cap.

Ready to put verified AI to work?

Ask a real question about your own area and watch it come back as a form, an analysis or a dashboard — checked before it publishes.