India Public Finance Portal
Research-grade Indian state budget data extracted from expenditure PDFs, with source provenance, validation checks, and multi-language worked examples
github.comIndia Public Finance Portal is an open research project by XKDR Forum that extracts and standardizes granular state budget data from official expenditure-volume PDFs into machine-readable CSV and NDJSON formats.
Rather than aggregating or flattening away inconsistencies, the repository retains source document hierarchy, line-item metadata, and row roles so researchers can audit any number back to the original budget page.
- Current Coverage: Karnataka state budget expenditure data spanning 9 fiscal years (2018–19 to 2026–27) across 7 annual volumes, and Tamil Nadu budget data covering 13 fiscal years (2014–15 to 2026–27).
- Auditability & Integrity: Each state package includes data dictionaries, demand-number-to-department lookup tables, machine-readable validation summaries testing internal summation rules, and cross-year accuracy metrics.
- Worked Examples Compendium: Fully reproducible analyses implemented across Python (standard library only), base R, and DuckDB SQL, packaged with a pinned Docker environment for one-command replication.
Details
- Kind
- Data source
- Run by
- Non-profit
- Topics
- Where
- India, Karnataka, Tamil Nadu
- Licence
- Open licence
- Cost
- Free
- Licence terms
- MIT
People
- Ajay Shah
- Susan Thomas