Customer Guide
This guide is for recruiting teams, customer operators, and tenant admins evaluating or running Veref in a hiring workflow.
What Veref does
Veref helps recruiters collect candidate-approved reference evidence, coordinate referee outreach, review trust signals, and package evidence-backed reports for hiring decisions.
The current product foundation centers on the reference workflow:
- A recruiter or approved agent starts a reference request.
- The candidate grants permission and supplies or confirms referee contacts.
- Referees complete structured reference conversations through public links.
- Veref summarizes evidence, risk signals, activity, and next actions.
- The recruiter approves and shares the final report.
Who uses it
| Actor | What they do |
|---|---|
| Recruiter | Starts requests, monitors progress, reviews evidence, approves reports |
| Coordinator | Helps move tenant workflows forward with lower authority than admins or owners |
| Candidate | Grants permission and confirms work history or referee contacts |
| Referee | Responds to reference requests through a public flow |
| Agent | Starts, inspects, or advances workflows through approved API or CLI access |
Read Product Concepts for the full actor and entity model.
Primary workflow
The default customer workflow is candidate intake:
- Recruiter creates or imports a candidate.
- Recruiter starts a reference request.
- Candidate receives an intake link and grants permission.
- Candidate submits referee contacts.
- Referees receive email or SMS outreach.
- Recruiter reviews the completed evidence package.
Use manual references when the recruiter or an approved system already has referee details and wants outreach to begin directly.
Consent and reuse
Veref treats candidate permission as a product gate. Reusing existing reference details or Passport evidence requires a candidate approval flow, not a silent import.
See Reference Reuse Permissions for the consent model and audit behavior.
Customer-safe outputs
Reports should be evidence-backed and recruiter-approved before sharing. Customer-facing material should avoid unsupported AI copy, internal route names, local commands, and implementation details.
For implementation boundaries, use Documentation Boundaries.