This guide covers how to configure Leaf as the data infrastructure for a farm management information system. Leaf fetches and standardizes field boundaries, machine files, and field operations from multiple machinery brands and provides satellite imagery, weather data, and grower connectivity tools.
Account structure
Use a single API owner per environment (production, test, staging). Each Leaf user represents one grower account in your FMIS.
This maps cleanly to how most FMIS platforms work:
- Each Leaf user corresponds to your internal representation of a customer.
- Adding a new grower means creating a new Leaf user.
- Usage and billing metrics are available per Leaf user, useful if you pass costs through to growers.
If your FMIS supports sub-accounts or partner access, you can create multiple API owners (e.g., one per region or partner). In practice, most FMIS integrations only need separate API owners for test vs. production environments.
Data needs
FMIS platforms typically consume several Leaf data products:
Field boundaries — synced automatically from connected providers. Leaf normalizes boundaries across John Deere, Climate FieldView, CNHi, and others into a consistent format.
Field operation summaries and images — planting, harvest, and application operations with field-level totals. Enable operationsImageCreation if you want map images generated for each operation.
Satellite imagery — in-season NDVI and other vegetation indices from Sentinel and Planet. Configured separately from machine data.
Weather data — historical and forecasted weather at the field or coordinate level.
Some FMIS platforms also need point-level data (standardGeoJSON or filteredGeoJSON) for custom map rendering or advanced analytics.
Processing rules
FMIS platforms generally want clean, ready-to-display data. The default configurations work well for most cases:
cleanupStandardGeojson (default: true) — leave enabled to get cleaned point data.
fieldOperationCreation (default: true) — leave enabled so Leaf automatically merges machine files into field operations.
operationsFilteredGeojson (default: true) — provides an additional filtered output with outliers and non-representative points removed. Useful for map rendering.
operationsImageCreation (default: false) — enable if you want Leaf to generate PNG images for each operation. Configure operationsImageAttributeCreation to control which attributes get images (e.g., yield, moisture, seed rate).
operationsImageAsGeoTiff (default: false) — enable if you need GeoTIFF format for integration with GIS tools.
unitMeasurement — set to IMPERIAL or METRIC for consistent units across providers.
operationsMergeRange and operationsMergeRangeHarvested — control how Leaf groups machine files into single operations. Defaults are 5 days for non-harvest and 21 days for harvest. Adjust if your growers’ operations span different time windows.
Sync granularity
FMIS platforms typically track multi-year field histories. Configure sync settings to match how far back your growers need data:
operationsProcessingRange — set this to cover the number of seasons your FMIS displays. If growers want 2-3 years of history, adjust accordingly.
customDataSync — consider enabling this to limit which fields Leaf fully processes, especially during initial onboarding when you’re connecting accounts with hundreds of fields.
organizationDataSync — relevant if you’re connecting growers with large John Deere accounts that span multiple organizations. Set to SELECTED_ONLY to process only the organizations you need.
Set up alerts for field events, machine file events, and field operation events. This lets your FMIS update in near real-time as new data flows through Leaf, rather than polling on a schedule.
What to read next
