GeoLens

GeoLens is an open-source, self-hosted catalog and map builder for geospatial data. It gives a team one place to search their datasets, preview them on a map, compose multi-layer visualizations, and serve the results through open standards, all running on infrastructure they control. The gap it fills is the one between scattered spatial files and a usable internal map service. Most organizations sit on shapefiles, GeoTIFFs, and GeoJSON spread across drives and buckets, with no shared way to find what exists or turn it into something people can actually look at.

You start by ingesting data directly. GeoLens accepts shapefiles, GeoTIFFs, GeoJSON, GeoPackages, Cloud-Optimized GeoTIFFs, and CSV, detects the coordinate reference system automatically, and builds a spatial index so the catalog stays fast even as it grows. It also handles VRT mosaics, which lets you stitch many raster tiles into one logical layer without pre-merging them by hand. Once data is in, it becomes a first-class catalog entry rather than a file someone has to remember the path to.

Finding the right layer is where the catalog earns its keep. Full-text search covers dataset names and descriptions, and when GeoLens is configured with vector embeddings you also get semantic search that matches on meaning rather than exact wording, so a search for flood zones can surface a layer named something less obvious. On top of that you can filter by geometry type across vector, raster, and table data, by CRS, and by date range, which narrows a large collection down to the handful of layers that actually matter for the task in front of you.

The map builder is where the datasets become a picture. It's a drag-and-drop, point-and-click composer for stacking multiple layers, editing styles, applying color ramps, setting filters, and drawing features by hand. It handles 3D terrain with hillshade and adjustable exaggeration, so elevation-heavy data reads properly instead of flattening out into a plain fill. Finished maps can be shared through signed links or dropped into another page with embed code, which turns an internal exploration into something a wider audience can view without needing an account on the system.

Underneath the interface, GeoLens is built on open standards rather than a proprietary format. It exposes OGC API endpoints for Features and Records with CQL2 filtering, and it's compatible with STAC and DCAT for cataloging. Because it speaks those standards, it drops into existing GIS workflows and connects cleanly to desktop tools like QGIS and ArcGIS. That matters for teams that already have a spatial stack and don't want a walled garden bolted onto the side of it, since the same layers they browse in GeoLens are reachable from the tools they already use.

It's aimed at GIS teams, data engineers, and organizations that work with spatial data and want to keep it in-house. Role-based access control separates admins, editors, and viewers, so an analyst can publish while a wider group only views. OAuth integration supports Google, Microsoft, and generic OIDC providers, which means it plugs into whatever identity system a company already runs instead of maintaining a separate login. Comprehensive audit logging tracks what happened across the catalog, the piece that lets a tool graduate from one analyst's machine to something a whole department can rely on without giving up governance.

What sets it apart is the combination of self-hosting and open standards in one package. Plenty of hosted mapping platforms will catalog and visualize spatial data for you, but they keep the data on their cloud and speak their own dialect. GeoLens runs entirely on your own infrastructure, collects no telemetry, and requires no external cloud account, so sensitive or regulated geospatial data never leaves your control. The catalog and the map builder ship together as one system rather than as separate products you have to stitch together and keep in sync.

On access, GeoLens is free and open source under the Apache 2.0 license. You deploy it yourself with Docker Compose for a straightforward single-host setup or Kubernetes for a larger one, and there's a community Helm chart for the Kubernetes path. That means no per-seat fees and no usage metering, with the tradeoff being that you run and maintain the service yourself. For a team that already operates its own servers and wants a spatial data catalog it fully owns, that tradeoff is usually the whole point. Teams that would rather not run it can still read the code, audit exactly what it does with their data, and adapt it to fit, which is the kind of freedom an open license buys even when raw convenience is not the goal.