DataFixer

DataFixer is a web app (with an API) that helps civic teams and journalists turn broken “open data” into analysis-ready tables. The brutal reality: many government portals publish CSVs with inconsistent schemas, missing metadata, weird encodings, and silent changes that break dashboards. DataFixer monitors selected datasets, detects schema drift, validates fields against rules you define, and produces a cleaned, versioned output with a changelog. It also generates a plain-English “what changed” summary and a data quality score so non-technical stakeholders can trust the numbers. The app is a combination traditional + AI: traditional pipelines do deterministic validation/transforms; AI is used only for mapping columns, suggesting fixes, and drafting human-readable release notes. You sell it to organizations that rely on open data but don’t have data engineers on staff.