FastOpenData is a one-person project working toward a single clean API over U.S. geographic, demographic, and address data. Today it can geocode an address. Joining that against Census, TIGER, and OpenStreetMap data is real, working ETL code in the repo — it's just not wired up to the live API yet.
# Geocode a US address (self-hosted Nominatim under the hood) curl -X POST http://<host>:8000/structured_geocode \ -H "Authorization: Bearer $TOKEN" \ -H "Content-Type: application/json" \ -d '{"street": "350 5th Ave", "city": "New York", "state": "NY"}'
FastOpenData{
"place_id": 127365481,
"licence": "Data © OpenStreetMap contributors, ODbL 1.0",
"osm_type": "way",
"lat": 40.7484, "lon": -73.9857,
"display_name": "350, 5th Avenue, Manhattan, New York, NY",
"category": "building",
"importance": 0.71
// this is a thin pass-through of Nominatim's response —
// no demographic/Census enrichment yet
}
No SDK, no schema, no dashboard is required for what currently works — just one endpoint.
Send a street/city/state, get back a Nominatim geocoding result (coordinates, OSM place data). That's the entire live surface area right now.
The repo already has working ingestion code for Census ACS/PUMS, TIGER/Line, OpenStreetMap, SIPP, AHS, Wikidata, and Foursquare Places — built for a graph engine, not yet exposed as address-lookup fields via this API.
The goal is still a single call that returns geocoding plus joined demographic/geographic fields. Getting there is the roadmap — it isn't done yet.
These are the actual datasets the ETL pipeline in this repo is built to ingest — none of them are joined into the address-lookup API today.
American Community Survey person and housing microdata, 1- and 5-year estimates.
Tract, county, state, PUMA, block-group, and address-point boundaries.
U.S. point-of-interest extract, updated via Geofabrik.
National housing survey data and value labels.
Survey of Income and Program Participation, person-level.
Geolocated entities and open places-of-interest data.
This is a solo project, developed in public. If you want to see where it's headed or follow along as it grows, the source is the most honest place to look.