What the source contains
The snapshot aggregates historical public Amazon hiring observations by canonical location. ZemGuru uses distinct Job IDs, first and latest retained dates, active calendar months, recent titles, employment metadata, postal areas and preserved pay when available.
How city content is composed
Content Engine v2 deduplicates recent rows by Job ID, derives narrowly bounded facts and selects reviewed English copy by content version, location ID and semantic slot. The result is reproducible: the same source and version produce the same page. Variation includes evidence headings and block order, while every rendered claim remains traceable to the snapshot.
What this pilot deliberately avoids
Repeated presence-period totals and average visibility duration are withheld from prominent claims while their creation and deduplication rules are being remediated. Raw state labels do not define geography; canonical region IDs do.
Historical does not mean open
A missing or old listing is not presented as a current vacancy. ZemGuru never emits JobPosting structured data for these archive pages. Visitors must verify current availability, requirements and application details on the official hiring source.
Why only nine cities
The generator can materialize a much larger catalogue, but the pilot publishes three cities in each supported country. A future city without enough factual coverage is automatically non-indexable and excluded from the sitemap.
Build reproducibility
The build manifest records source time, route list, content signature, factual coverage and content version. A validator checks every generated route, SEO signal, evidence block, selected pool, closest-page similarity, internal link, Umami configuration and ShiftSniper attribution before deployment.