I’m building a climate-tech MVP whose core value lies in smart geospatial analysis. The stack is already sketched out—Python on the server side, PostGIS for spatial storage, and Uber’s H3 for multi-resolution indexing—but I need an experienced backend engineer to turn that blueprint into a production-ready service. The workflow is centred on two data streams: recent satellite imagery (mostly raster metadata for now, not full pixel pipelines) and live GPS tracks from field devices. Your job is to design the schema, stand up the ETL routines, and implement performant spatial queries that combine both datasets for downstream models and client apps. Key expectations • Build a clean Python codebase (FastAPI is my default choice, but I’m open if you have a better idea) • Configure PostGIS and integrate H3 indexing so hex-level aggregations run effortlessly across zoom levels • Create ingestion scripts that pull satellite imagery metadata and GPS files, validate them, and load everything using async workers • Expose the processed outputs through well-documented REST endpoints, complete with unit tests and a Dockerfile for reproducibility When you reply, please link to past work that demonstrates heavy geospatial lifting—especially anything involving PostGIS tuning, H3, or large raster/vector blends. If you have open-source repos, conference talks, or published notebooks, even better. I’m moving quickly, so the first milestone will be a running container that ingests sample data, populates the database, and returns an aggregated hex grid via the API. Looking forward to seeing how you’ve solved similar mapping challenges in the past.