Earthobservation,onschedule.
A geospatial workflow platform for teams who build satellite-imagery pipelines — and for the agents that run them. Compose, schedule, inspect. Ship structured outputs, not screenshots.
Unified across the providers you already use
One surface for analysts
and the agents beside them.
The canvas and the API operate on the same workflow document. What you design by hand, an agent can produce in JSON — and vice versa.
Compose, schedule, inspect.
Versioned, reproducible, agent-readable.
Drag a step or POST a JSON spec — both go through the same validation and execution engine. Every run leaves a deterministic trail.
A shared canvas
Saved workflows, AOIs, and parameters — searchable across the team.
Time series, not stills
NDVI, NDWI, NDSI, and custom indices — emitted as series, not just rasters.
Unified credential layer
One billing relationship across Sentinel-2, Landsat-9, MODIS, and Planet — or bring your own keys.
Native MCP & REST
A workflow is a JSON document. Agents can create, run, and inspect them programmatically — no scraping the UI.
Questions, answered.
Which sensors and providers are supported?
Out of the box: Sentinel-2 (ESA), Landsat-9 (USGS/NASA), MODIS (NASA), and Planet Labs. Bring your own credentials, or use ours under a unified billing layer.
Can agents run workflows without the UI?
Yes. Every workflow is a versioned JSON document. Agents create, execute, and inspect runs through the REST API or our MCP server — the canvas is one of several surfaces, not the source of truth.
What do runs produce?
Structured artefacts: GeoJSON change polygons, NDVI and other index time series, scene-level statistics, and signed cloud-storage URLs for derived rasters. Everything is machine-readable.
How is data handled?
Encrypted in transit and at rest. We never train on customer data. Enterprise customers can pin processing to a region (US, EU) and bring their own object storage.
Is there an SLA?
Pro and Team run on shared capacity with best-effort throughput. Enterprise includes a 99.9% uptime SLA, reserved capacity, and a named support contact.
How does pricing scale?
Seat pricing is flat. Scene-level processing is metered separately on Pro and Team, and included in committed capacity on Enterprise. Talk to us if you're running at petabyte scale.
Run your first pipeline
before the next pass.
Provision a workspace, point at an AOI, and ship structured outputs to your agents.