World Bank Open Data

World Bank Open Data is a free portal of global development statistics covering economies, populations, health, education, infrastructure, poverty, trade, and the environment. Its flagship is the World Development Indicators (WDI), a compilation of more than 900 indicators with time series for 210 countries stretching back to 1960. Because the figures are assembled to be internationally comparable, you can line up the same measure across countries and across decades on a like-for-like basis, which is what makes it useful rather than just large.

Beyond the headline indicators, the portal hosts a Microdata Library of underlying household and firm survey data, a Data Catalog for bulk downloads, and thematic databases on debt, finance, and specific development programmes. The main REST API returns indicator data as JSON or XML and is straightforward to page through by country and year; there are also well-maintained community wrappers (for example the wbdata and world_bank_data packages in Python and R) that save you writing the request logic yourself.

The thing to plan for with RAG is that this is numeric time-series data, not prose, and a language model cannot retrieve or reason over a wall of numbers well. The reliable pattern is to turn the series you need into short natural-language statements (for example, "In 2020, GDP per capita in Kenya was X current US dollars") and embed those, or to keep the numbers in a database and let the model call it, rather than dumping raw tables into a vector index. It suits economics assistants, development research tools, and any application that needs authoritative country-level indicators to ground its answers.

Watch for gaps and revisions: not every country reports every indicator every year, values are updated retrospectively, and definitions change over time, so cite the vintage of the data you used. The data is released under CC BY 4.0, so commercial use is fine as long as you credit the World Bank. For deeper regional coverage, pair it with Eurostat, the IMF, or OECD statistics, which share a similar structure.