I used to work on a delivery area feature where I used Shapely to simplify and dissolve subsets of a shapefile. I thought that for this simple task, I don't need to use Python scripts or host a WFS/WPS

Inspired by how Cloudinary and imgproxy transform images through URLs, I built Geoblob (geoblob.com) to do the same thing but for geospatial data.

A few examples (simplified for readability - actual URLs include a signature and API key):

Filter features from a dataset: https://api.geoblob.com/{signature}/{api_key_id}/@dataset:bo...

Simplify and dissolve: https://api.geoblob.com/{signature}/{api_key_id}/@dataset:bo...

Get bounding box, area, and centroid without downloading anything: https://api.geoblob.com/{signature}/{api_key_id}/@dataset:bo...

You can change the file extension to change the output format - .geojson, .kml, .gml, .svg, .wkt, .mvt.

Operations include buffer, clip, convex hull, Voronoi, K-Means clustering, hex/square grids, and more.

Would love to hear your thoughts - is this something GIS people would actually use? What operations or formats would be useful?