geomermaids · Frequently asked questions
What Geomermaids is, the cloud-native geospatial work it does, the formats and tools behind it, and how to get started.
What is Geomermaids?
Geomermaids is a cloud-native geospatial consultancy run by Guillaume Sueur from the Boston, Massachusetts area. It designs and builds geospatial systems (data pipelines, APIs, dashboards, and machine learning workflows) using open source wherever it fits, so spatial data is cheap to query, easy to reuse, and independent of any single vendor.
Who is behind Geomermaids?
Guillaume Sueur, founder of Geomermaids LLC. He started out contributing to national and regional Spatial Data Infrastructure (SDI) projects in France and has spent the last decade on cloud-native geospatial formats and the tooling around them, with more than 25 years in GIS overall.
What is cloud-native geospatial?
It is an approach where geospatial data lives on object storage (S3, R2) in formats designed to be read directly over HTTP, in pieces, without a running server: Cloud Optimized GeoTIFF for rasters, GeoParquet for vectors, and STAC to catalog it. Queries pull only the bytes they need, which gives cheap, parallel, vendor-neutral access at scale.
What services does Geomermaids offer?
Geospatial data management, data publication and visualization, GeoCloud architecture, machine learning for geospatial, earth observation and imagery (Cloud Optimized GeoTIFF, STAC catalogs, on-demand tile services), and migration off proprietary GIS stacks. Engagements are tailored consulting, delivered remotely to teams worldwide.
Which geospatial formats and tools do you work with?
Mainly Python and SQL, with deep PostgreSQL/PostGIS experience and a recent focus on DuckDB, to which Guillaume has contributed patches. The data layer is usually object storage (S3, R2) with GeoParquet, Cloud Optimized GeoTIFF, and STAC. Compute and orchestration are chosen to fit the problem rather than forced into one stack.
Can you migrate us off Esri/ArcGIS or another proprietary GIS?
Yes. Open source migration is a core service: replacing closed GIS stacks with PostGIS, QGIS, and cloud-native equivalents without losing capability. The goal is to keep your workflows intact while removing per-seat licensing and vendor lock-in.
Do you work remotely, and where are you based?
Geomermaids is based in the Boston, Massachusetts area and works remotely with teams worldwide.
What is GeoParquet and why use it?
GeoParquet stores vector geospatial data in the columnar Parquet format, so it can be queried directly from a URL or object storage with engines like DuckDB, with no database server required. It is compact, fast to scan, and lets a query prune to only the columns and row groups it needs. See the GeoParquet cookbook.
What is a Cloud Optimized GeoTIFF (COG)?
A COG is a GeoTIFF laid out so a client can read just the tiles and overviews it needs over HTTP range requests, instead of downloading the whole file. It powers on-demand tiling and analysis straight from object storage. See the COG cookbook.
Do you do GeoAI or machine learning on geospatial data?
Yes. Geomermaids builds machine learning workflows for geospatial problems using open source frameworks, from feature extraction on satellite and aerial imagery to training and inference pipelines that run over cloud-native data.
How can I work with Geomermaids?
Send a note through the contact form on the homepage with your project, dataset, or architecture question, and Guillaume will be in touch.