Methodology

How this data is built

Everything public on Zavmo comes from a careers graph with named sources, honest labels, and human gates. This page is the whole method, short enough to actually read.

The sources

  • Regulated qualifications come from the Ofqual Register of Regulated Qualifications, used under the Open Government Licence v3.0, including awarding body, regulated framework (RQF) level, and registry identifiers.
  • Skills are the UK's National Occupational Standards (NOS), the published national definitions of competent performance.
  • Roles are synthesised job profiles connected to the standards and qualifications above.
  • Labour-market figures (pay, demand) appear only when they trace to a nameable authoritative source. A figure we can't source is not shown. There are no placeholder numbers anywhere on this site.

Authoritative vs suggested

Role → skill links come from the standards themselves and are treated as authoritative. Skill → qualification and role → qualification links are machine-inferred (vector similarity, top matches per standard). Everywhere they appear, on pages, widgets and reports, they are labelled "Suggested match" and presented as a starting point for a human decision, never as regulator fact.

The publish gates

  • The value test: a page is published to search only if it would deserve to exist if Google did not: genuine synthesis, not reprinted regulatory data.
  • The two-state rule: pages without enough substance stay public and human-usable but are excluded from search indexing until the data earns it.
  • The editorial gate: data reports are generated as dated snapshots from a read-only audit of the graph, then reviewed by a human before publication. Each report names the audit it came from.

Corrections

Found something wrong? Email hello@zavmo.ai. Corrections ship as data fixes, not quiet edits, and reports carry their as-of date so a corrected edition is distinguishable from the original.