NYC Open Data

NYC Open Data is the City of New York's open data portal, run by the Open Data Team in the Office of Data Strategy under the city's Open Data Law of 2012, which requires city-agency data to be open by default. Each agency has an Open Data Coordinator who publishes and maintains its datasets, so the catalogue spans the whole of city government: 311 service requests, restaurant inspections, taxi trips, building permits and property records, street works, tree plantings, birth records and much more.

It runs on the Socrata platform. Every dataset can be previewed and exported from the web as CSV or JSON, and queried programmatically through the Socrata Open Data API (SODA); a free application token lifts the rate limits for heavier use, so you can pull just the slice you need rather than a whole dataset.

Most of NYC Open Data is tabular, and a table of numbers is not a ready-made RAG corpus. The parts worth retrieving over are the text-rich ones: the free-text complaint descriptions in 311 service requests, inspection findings, and the agency reports and documentation attached to datasets. As with any statistical source, turn rows into short natural-language statements before you embed them, so the retriever has readable text to match against rather than bare cells. Think of it as a civic-data layer for a New York City assistant, a policy or operations research tool, or a local chatbot.

The data is public and free to reuse, including commercially, under the portal's Terms of Use, which ask that you do not misrepresent it and note that it is provided as is. Datasets refresh on their own schedules, from daily to annual, so check the update date on anything time-sensitive.

Among its neighbours, data.gov is the United States national catalogue when you need coverage beyond the city, and the London Datastore is the equivalent for London. Reach for NYC Open Data when your questions are specifically about New York City.