Hey Product Hunt 👋 I built Open Economics because working with Brazilian official economic data was much harder than it should be. The data is public and high quality, but every institution exposes it differently. BCB, IBGE, Tesouro, SICONFI, Comex Stat, MTE, ANP, EPE and CVM all have different APIs, identifiers, dimensions and semantics. I wanted one consistent layer that developers, analysts and AI agents could use without losing the meaning or provenance of the original data. Open Economics is now: • free and open source • no signup or API key required • read-only • accessible through REST • available as a remote MCP server with 19 tools • built around official Brazilian data sources • designed to preserve dataset identity, dates, units, metadata and provenance One of the hardest problems wasn’t combining APIs — it was semantic correctness. A query like “Selic” can refer to multiple legitimate official series. Open Economics separates concept resolution from data availability so it can identify the intended official dataset instead of silently returning the closest-looking series. I also found an interesting interoperability issue while testing the MCP server: some clients could call the tools successfully but weren’t exposing structuredContent to the model. I changed the server so important values, dates, units and provenance are available in both structured and model-visible output. The project is live today and I’m still improving the coverage. I’d really like feedback from developers, economists, researchers and people building AI agents: What Brazilian economic-data query is still unnecessarily difficult for you today?

Open Economics

Brazil’s official economic data, ready for APIs and AI

Brazil’s official economic data, ready for APIs and AI