Methodology
Folio Australia is a Public Records Intelligence Platform. It does not publish opinions, verdicts, scores or narratives. Every piece of information visible on the site comes directly from an Australian public record.
Methodology disclaimer
Folio shows public-record connections only. It does not allege wrongdoing, corruption or improper influence.
Every fact on this page is drawn from an original Australian public record and links back to its source. Nothing is scored, inferred or AI-summarised.
- Source-backed only. Every fact links to its original document.
- No AI-generated conclusions. No corruption scoring. No political framing.
- No inferred relationships. Two entities are only connected when a public record names them together.
- Uncertain matches are held in a human review queue and never merged automatically.
- Records with a stable identifier (an ABN) can be linked across datasets safely; everything else requires human confirmation.
Each configured data source runs through the same pipeline:
- The source is fetched from its public URL.
- A parser converts the body into structured records.
- Records are deduplicated by content hash so re-runs are safe.
- Each record is stored raw in a snapshot table and then upserted into its canonical table.
- The record is mirrored into the entity graph as entities, source-backed relationships and timeline events.
- The outcome is recorded on the source's health status.
When an ABN is present, it acts as a globally unique identifier and records naming the same ABN are automatically linked to the same organisation entity. Without an ABN, entities are keyed by a normalised name within the dataset that discovered them, and any proposed cross-dataset match is routed to the match queue for human review.
If you notice an error, the fix is upstream: contact the original publisher of the record. Once they update their record, Folio Australia will re-ingest and reflect the change on the next run.