Data Methodology

From public records to structured signals

How raw public data becomes something an agent can reason about.

Ingest public records such as company filings, planning portals and official notices.

Map each record to a consistent schema: entity, event type, location and date.

Attach the original source, retrieval time and event date to every record.

Assign a confidence level and, separately, an inferred opportunity score.

Every signal links back to the public record it came from, so agents and people can check it.

We separate when something happened from when we found it, so recency is never ambiguous.

High, medium or low, based on how direct and complete the source is.

Planned coverage spans England, Scotland, Wales and Northern Ireland, at town and region level.

Facts vs. inferences

Every response keeps what a source states apart from what we conclude from it.

Verified fact

"A planning application was submitted for change of use from office to light industrial on 6 October."

Directly stated in the source record. Marked inferred: false.

Inferred opportunity

"The occupier may need fit-out contractors and a commercial energy supplier."

Our interpretation, with a score and rationale. Marked inferred: true.