Skip to content
OpenPalm AI
DocumentationOpen the platform
Menu

Getting started

Explore your first geospatial workflow

Select the real Selangor study, inspect satellite vegetation and predicted canopy height, then download the scientific run. Saved research is available without model credentials.

Evidence boundary

Real study and demo fixtures are separate. SAM segmentation is not implemented; synthetic parcel boundaries are never presented as model predictions.

  1. Select an area of interest

    Open Geo Intelligence. The Real study view uses a pinned 19.6704 ha Selangor research window, not a verified plantation boundary.

    Open Geo Intelligence

  2. Explore the real analysis

    Toggle QA-masked Sentinel-2 NDVI and Meta-predicted CHMv2. Check the historical acquisition dates: the sources are 589 days apart, so they do not describe simultaneous conditions.

  3. Inspect a pixel

    Click within the preview to inspect its value and units. NDVI is a vegetation index; CHMv2 is a model prediction, not ground-truth height or individual-tree detection.

  4. Export the scientific run

    Choose Download run bundle for the analysis, source metadata, provenance and saved research. To explore synthetic parcels and GeoJSON instead, switch to Demo fixtures.

# Set OPENPALM_URL to this deployment's origin. No API key required.
curl --fail --output study-run.zip \
  "$OPENPALM_URL/analysis/run-f6c78baf04ed06bf81eda7d6/bundle-with-research-42c27339c57324ba.zip"