Where the data comes from

AuditGava aggregates what Kenyan government agencies already publish. Most figures here trace to a named document; where one does not, the page carrying it says so and names the method instead. No private sources, no opinion.

Documents indexed

Publishing agencies

The freshness age below shows how recently we last reached each agency. Several are months behind; the site does not currently update every source nightly, whatever a headline elsewhere may suggest.

Data health

Reading how many rows each dataset holds and how long since any of them changed.

How this works

Our ETL pipeline fetches PDFs and spreadsheets from each agency's official portal, extracts line-items using table-extraction tools and rule-based parsers, and writes them to a canonical schema.

Coverage of the evidence trail is uneven, and we would rather say so than imply otherwise. Audit findings and national debt instruments carry a provenance pointer — a source document and, for audit findings, a page reference — and can be checked through the verification endpoint. County budget lines and the historical debt timeline cannot yet: they are not wired into that endpoint. County figures come from two different places, and each page says which: the headline budget, spend and execution rate are read from the Controller of Budget's implementation review where that parse landed, while the per-sector split and any future-year period are modelled from the CRA equitable-share formula. Pages carrying modelled figures say so on the page.