Skip to main content
This documentation covers Public Records — criminal record compliance and reportability for Consumer Reporting Agencies and background screening companies. You submit a criminal record (XML or mapped JSON) with candidate information. SNH AI returns a synchronous decision for every charge — whether it can appear on a background report, why, and where to route the work. Both are SNH AI Background Check products. Use this site only for criminal Public Records.

What it does

You send a criminal record and information about the person being screened. The service returns a decision for every charge:
  • Reportable — the charge can be included in the background report
  • Not Reportable — compliance rules prevent reporting (e.g. too old, minor traffic, aged dismissal)
  • Manual Review — a human must decide (missing data or unresolved rule)
Along with each decision, you get:
  • A plain-language rationale
  • Legal citations (e.g. 15 U.S.C. § 1681c(a)(5))
  • An identity match score (candidate vs record subject)
  • A routing recommendationAutomation* (no human review needed), Auditor* (human review recommended, optional), or Insufficient Data* (human review required)
*See Glossary → Routing Queues. Jurisdictions: Evaluation applies federal rules, state rule packs for all 50 US states, the District of Columbia, and US territories with packs (AS, GU, MP, PR, VI), plus local (county) context from the record (jurisdiction_state / jurisdiction text). Your organization may also have client- or tenant-specific rules from onboarding.

How it works

  1. Create your organization profile and receive API credentials (Account Setup)
  2. Authenticate with a Bearer JWT (Authentication)
  3. Evaluate a record via POST /evaluate and process the sync decision (Quickstart)
One evaluate call returns the full decision — no webhooks or polling. Each call evaluates one search. See How It Works, plus Error Handling and Retry Logic for failures.

Common use cases

  • Automate federal / state / local compliance decisions for CRA and screening workflows
  • Auto-clear clearly not-reportable charges to cut manual review
  • Route reportable / unclear charges to auditor or researcher queues with rationale and citations
See Examples for request patterns and Glossary for shared terms.

Support