.png)
Leaf builds the API that standardizes fragmented agricultural data from platforms like John Deere Ops Center and Climate FieldView, serving enterprise customers including Syngenta and Bayer. Their mission: make agricultural data accessible at scale—enabling analysis across millions of acres.
Leaf needed to visualize large agricultural datasets for development and customer support, but desktop GIS tools weren't cutting it. Performance issues meant crashes on datasets over 1GB—a non-starter for analyzing a million acres at once. Data prep required extensive scripting just to import CSVs with embedded coordinates. And collaboration meant screen-sharing calls or static screenshots with no way for stakeholders to explore data independently.
Leaf's new workflow is simple: export data as CSV, GeoJSON, or Parquet, then drag-and-drop into Felt. The "aha moment" came when the team discovered Felt could parse CSVs with embedded geo-data directly—no scripting required.
Instead of being bottlenecked by in-person screen shares, the team now distributes maps with shareable links, viewable from any device. With data flowing from S3, MongoDB, and Postgres, Leaf has a unified analytical environment on AWS. Using Felt AI which is powered by Claude Opus 4.6 they’re prototyping customer-facing analytics widgets without engineering support.
Looking ahead, Leaf is exploring Felt's new Wherobots integration to combine large scale geo processing and analysis with the power of building maps, apps and dashboards in seconds.
