Enterprise AI adoption.
AI-vendor relationships can surface in public subprocessor disclosures, governance documentation, and technical references. These sources can help identify adoption while preserving the distinction between direct use, an integration, and use through another provider.
Coexistence is central to this category. A company can add a model provider while continuing to use another. Our research treats that as its own event.
Applicant tracking systems.
Company career sites and public job-board endpoints can expose the recruiting platform behind a hiring workflow. Following the same company over time can reveal changes worth investigating.
A new job board can belong to a subsidiary, a regional hiring team, or a parallel recruiting workflow. Those possibilities need to be checked before labeling a migration.
Coverage must be measured, not implied.
- Company identity and the business unit represented by the source.
- Vendor category and the particular product observed.
- Observation history, collection gaps, and source availability.
- Geographic and company-size biases in the observable sample.
Expand only where evidence supports it.
Customer support, identity, data infrastructure, and other back-office categories require their own source evaluation and matching rules. Evidence from one category cannot establish coverage in another.
The goal is a useful, inspectable record within a defined scope. Breadth comes after a repeatable evidence standard.