Credentialing teams and specialist recruiters have the same recurring job: a public register of licences, permits or registrations matters to their work, and it changes without telling anyone. The usual method is to re-read the current export and hope something looks different from last quarter, which finds additions occasionally and removals almost never.
Comparing two dated exports finds all three kinds of change, and produces a record of having looked.
Every record carries a stable id and the register's source URL, so each reported change points back at the original public entry rather than at a summary of it.
A removal is a question, not an answer. A record can leave an export because the export was filtered differently, because the register was partially unavailable when it was taken, because an identifier changed, or because of a routine administrative correction. Open the source record and confirm before acting.
It compares snapshots the customer is authorized to supply — each input must declare authorized: true, a register name, and a source URL. It does not scrape registers, build profiles of people, aggregate sources, rank or score anyone, decide suitability, or assert that a person is compliant, safe, or eligible. A public register entry is a fact about a record; it is not a judgement about a person, and a diff of two exports must not be used as one.
Export the register on a regular cadence, keep each export as a dated snapshot, and compare consecutive snapshots rather than re-reading the current one. A comparison gives you the three things a manual review does not reliably produce: the records that appeared, the records that disappeared, and the fields that changed on records that stayed. Keep the source URL with each record so a reviewer can open the original entry instead of trusting the summary.
No, and treating it that way is the main risk in this workflow. A record can vanish from an export because the export was filtered differently, because the register was partially unavailable, because an identifier changed, or because of an ordinary administrative correction. A removal is an observation that tells you where to look. Open the source record and confirm before acting on it, and never let a diff become the basis of a decision about a person on its own.
No. It compares two snapshots that you are authorized to supply and reports what differs between them. It does not scrape registers, aggregate sources into profiles, rank or score people, judge suitability, or state that anyone is compliant, safe, or eligible. Each input must be marked as authorized and must carry the register name and source URL, so every reported change points back at a public record a reviewer can read.