Skip to Content
The GeoInsight Grid

Introduction

GeoInsight turns Earth observation and geospatial data into a unified vocabulary of Spatial Tokens, so applications and AI models can reason about geography the same way language models reason about text.

ie-bar

At its core, GeoInsight is agnostic to the data it harmonizes. Rasters, vectors, point clouds, tabular data - satellite data, sensor, IoT, or your own datasets - are all mapped into the same DGGS (Discrete Global Grid System) representation. You don’t need a different integration for every data type or provider; anything with a spatial footprint can be tokenized and queried the same way.

Why Spatial Tokens

Most geospatial systems organize data around projections and tiling schemes built for maps, not computation, and every new data source tends to mean a new ingestion path, a new schema, a new set of assumptions. GeoInsight instead maps everything onto a discrete global grid, so any observation (regardless of source, sensor, or original format) gets a stable cell address (a Spatial Token) on the same global grid.

This gives you:

  • Data-source agnosticism. Whether it’s satellite imagery, vessel tracking data, vector layers, or your own tabular data, it’s tokenized into the same grid and queried through the same interface.
  • Consistent spatial units. Tokens from different sources and sensors are directly comparable, with local tokens nesting naturally into regional and continental context.
  • Standards-based access. The API implements the DGGS OGC API  standard, so you’re working against an open specification, not a proprietary format.
  • Scale invariance. The same token hierarchy works from street level to continental scale, with no projection distortion if the grid is equal-area
  • Sovereign, EU-based infrastructure. Data is hosted and processed within the EU, which matters if data residency and sovereignty are requirements for your organization.

GeoInsight’s DGGS engine is powered by GeoPlegma , our open-source DGGS library (Rust core with JavaScript bindings). The core tokenization and grid infrastructure is open source; the analytics and model layer built on top of it is GeoInsight’s proprietary product.

Harmonization & interoperability

Different data sources rarely agree on format, projection, resolution, or update cadence - which is usually where integration work actually goes. GeoInsight’s tokenization step is also a harmonization step: every source is normalized into the same DGGS cell structure, so a raster from one provider and a vector layer from another end up addressable and comparable in exactly the same way.

Because that harmonization layer is built on the open DGGS OGC API standard rather than a proprietary format, GeoInsight is also interoperable by default:

  • Cross-source analysis is native, not custom. You don’t write source-specific reconciliation logic, i.e. two datasets tokenized into GeoInsight are already aligned to the same grid and can be queried and combined directly.
  • Standards compliance means portability. Because the API follows an open OGC specification, GeoInsight fits into existing geospatial toolchains and pipelines rather than locking you into a bespoke integration.
  • New sources don’t break existing ones. Adding a new data source extends the same token space rather than requiring a parallel access pattern, so integrations you’ve already built keep working as coverage grows.

What data can be tokenized

GeoInsight is built to ingest and harmonize any geospatial data source, including:

  • Optical and SAR satellite imagery (e.g. Sentinel-1/2)
  • Maritime tracking data (AIS)
  • Vector datasets and point clouds
  • Thermal and LiDAR data
  • IoT sensors
  • Your own tabular or custom spatial datasets
  • Anything the contains location.

Whatever the source, it’s mapped into the same Spatial Token space and exposed through the same DGGS-indexed API, which means that adding a new data source to your analysis doesn’t mean learning a new access pattern.

How the API is structured

Access to GeoInsight is organized in tiers, so you can start simple and grow into more advanced capabilities as your use case matures:

  1. Data API access: query harmonized, DGGS-indexed raw data directly.
  2. Derived products: calculated indices and pre-computed spatial products built on top of the raw data.
  3. AI analytics: advanced, model-driven analysis over the harmonized data layer.
  4. Enterprise / on-premises: custom deployments for organizations with dedicated infrastructure needs.

Most developers will start with tier 1, the core data API, which is what the rest of this documentation focuses on first.

Ways to access GeoInsight

  • REST API: the primary way  to query the harmonized data layer directly.
  • MCP server: GeoInsight also exposes its data API through an MCP  (Model Context Protocol) server, so AI agents and LLM-based tools can query GeoInsight data natively.

Next steps

Quickstart