> ## Documentation Index > Fetch the complete documentation index at: https://docs.pr.snh-ai.com/llms.txt > Use this file to discover all available pages before exploring further. # Glossary > Background screening and API terminology explained Key terms used throughout this documentation and in API responses. ## Platform Terms | Term | Definition | | ---------------- | ------------------------------------------------------------------------------------------------------------------------------------ | | **Organization** | The company using the platform. An organization has its own API credentials, environment access, and service configuration. | | **User** | An individual accessing the platform on behalf of an organization. | | **API Key** | Authentication credential issued per organization, per environment. Used to obtain JWT tokens. Format: `eval-api-key-`. | | **Request** | A single API submission to any endpoint. | | **Search** | A customer-initiated lookup or evaluation workflow, identified by a `search_id`. | | **Result** | The completed output from a service, returned in the API response. | ## Evaluation Decisions Reportability labels (`REPORTABLE` / `NOT_REPORTABLE` / `MANUAL_REVIEW`) appear at every level of the decision hierarchy: `charge_decision` on each offense, plus aggregated `case_decision`, `court_decision`, and `record_decision` on parent nodes. Routing queues (`Automation` / `Auditor` / `Insufficient Data`) are separate from decision labels and appear as `routing.queue` on offenses, plus aggregated `case_queue`, `court_queue`, and `search_queue`. | Term | Definition | | ------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Reportable** (`REPORTABLE`) | Charge can legally be included in a background check report under applicable federal and state laws. | | **Not Reportable** (`NOT_REPORTABLE`) | Charge filtered out by compliance rules — cannot be reported (e.g., beyond lookback period, sealed, expunged). | | **Manual Review** (`MANUAL_REVIEW`) | Confidence is low or no compliance rule matched — a human must make the reporting decision. | | **Queue → Decision mapping** | At aggregated levels, `*_decision` is derived from `*_queue`: `Automation` → `NOT_REPORTABLE`, `Auditor` → `REPORTABLE`, `Insufficient Data` → `MANUAL_REVIEW`. | ## Routing Queues | Term | Definition | | --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Automation** | Charge can be fully automated with no human review. Assigned when the offense is **not reportable**, or when the identity **does not match** the candidate. | | **Auditor** | Charge requires manual audit/verification. Assigned when the offense is **reportable** and the identity **matches** the candidate. | | **Insufficient Data** | Charge lacks enough information to make a determination (e.g., missing disposition, missing identity fields). | | **Queue Rollup** | `case_queue`, `court_queue`, and `search_queue` are computed from per-charge `routing.queue` values using the precedence `Insufficient Data` > `Auditor` > `Automation`. | ## Criminal Record Terms | Term | Definition | | --------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **candidate\_info (request)** | Required top-level object on `record` for both XML and JSON paths. Must be non-empty (not `null`, missing, or `{}`) with non-empty `first_name`, `last_name`, and `date_of_birth`. Optional: `middle_name`, `ssn`, `address`. Must **not** appear inside `record_json`. | | **submission\_type (request)** | Top-level field on `record`. Required `"resubmit"` when `record_json` is provided. Omit on initial XML submissions. Must **not** appear inside `record_json` or `record_json.search`. | | **order\_id / order\_number (request)** | Required top-level fields on `record` alongside `search_id` and `search_date`. Must be non-empty strings — not `null`, missing, or `""`. Must **not** appear inside `record_json` or `record_json.search`. | | **cases (request)** | **Optional** array of `{ court_search_id, offense_id }` pairs on `record`. Omit or send `[]` to evaluate all offenses (count-reconciliation gate not applied). When non-empty, must enumerate **every** court and offense — distinct court search ID count and total offense ID row count must exactly match parsed XML/record\_json. Multiple rows may share the same `court_search_id`. Partial/filtered subsets are rejected (XML-Transformer HTTP **422** internally; `/evaluate` returns HTTP **500** / `failed`). | | **Offense** | A single criminal charge within a case. A case can contain multiple offenses. | | **Case** | A court case containing one or more offenses. | | **Disposition** | The outcome of a criminal charge (e.g., Guilty, Dismissed, Deferred). | | **Charge** | The specific criminal charge description (e.g., "THEFT OF PROPERTY"). | | **Type Class** | Offense severity: `"FELONY"`, `"MISDEMEANOR"`, or `"INFRACTION"`. | ## Compliance Terms | Term | Definition | | --------------------- | ------------------------------------------------------------------------------------------------------------------------- | | **FCRA** | Fair Credit Reporting Act — the federal law governing background checks. | | **7-Year Lookback** | Under FCRA, most non-conviction records older than 7 years generally cannot be reported. Some states have stricter rules. | | **Lookback Period** | The maximum age of a record that can be reported. Varies by state and offense type. | | **Sealed / Expunged** | Records that have been legally sealed or expunged and are not reportable. | ## Identity Matching | Term | Definition | | --------------- | --------------------------------------------------------------------------------------------- | | **Match Score** | Overall weighted score (0–1) indicating how closely the candidate matches the record subject. | | **isMatch** | `true` when the match score meets the threshold for a positive identity match. | ## API Terms | Term | Definition | | ------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Search ID** | Unique identifier for a single background check search. | | **Correlation ID** | Unique UUID generated by SNH AI for each request. Returned in both `data.correlation_id` and `meta.correlation_id`. Include this when contacting support. |